Method and system for optimization of general problems

[0001] This application is a continuation of U.S. application Ser. No. 10/745,394, filed Dec. 23, 2003, now U.S. Pat. No. 7,072,723, which is a continuation-in-part application of U.S. application Ser. No. 10/693,729, filed Oct. 23, 2003, which claims the benefit of provisional patent application Nos. 60/420,921 and 60/420,922, both filed Oct. 23, 2002.

[0002] The present invention is related to general optimization methods employed to determine relatively optimal solutions for various types of problems and to control of computational processes and construction of computational approaches to solving complex problems and, in particular, to a general computational method and system for minimizing a general class of objective functions, for adjusting solutions to optimization over time, and for constructing and achieving relatively optimal computational control over a wide range of computational processes.

[0003] Significant research-and-development effort is currently being applied to the mathematical and computational field of optimization. Optimization techniques are applied to a wide variety of everyday and complex theoretical problems in order to find solutions to those problems, including cost-effective and time-efficient control strategies for complex systems and organizations. For example, optimization techniques may be applied to schedule labor and equipment resources within a large manufacturing organization in order to most cost-effectively produce a given amount of manufactured product to meet a projected demand for the product at various points in time. As another example, optimization techniques may be applied to a large control problem, such as the control of traffic lights within a city, in order to allow for as large a volume of traffic to move through the city as possible without generating unreasonable delays at various intersections and for those traveling on various secondary and tertiary routes within the city. Many natural systems can be viewed as seeking relatively optimal solutions to complexly defined problems governed by physical characteristics and principles and constrained by a large number of physical constraints. For example, complex chemical reactions can be considered to be governed by rather simple, thermodynamic principles by which individual components, such as molecules, seek minimum energy states and maximum entropy, constrained by various ways in which molecules can reconfigure themselves and exchange energy amongst themselves according to quantum mechanics. Even more complex biological systems can also be considered to be governed by chemical and thermodynamic principles as well as by analogous principles involving information content and exchange, compartmentalization and modular design, and other considerations. Thus, optimization problems may encompass an extremely wide variety of mathematically and computationally expressed models of natural phenomena, design, organization, and control of complex systems, and organization, transmission, and processing of information.

[0004] Current approaches for finding near optimal and optimal solutions for mathematically modeled problems are limited. When the number of decision variables and constraints in such problems increases from the small number of variables and constraints normally employed in simple, textbook problems to the large number of variables and constraints normally encountered in real-world systems, the computational resources required for seeking near optimal and optimal solutions increase exponentially in most cases. Current techniques cannot be satisfactorily applied to any but the simplest types of problems. Many currently available techniques involve applying oppressive constraints to optimization models, such as requiring variables to be continuous and requiring the hyper-dimensional volume representing the set of possible solutions of optimizations problems, or problem domain, to be convex.

[0005] Researchers, developers, system modelers, and investigators of many different types of complex system behavior have therefore recognized the need for more generally applicable and efficient methods and systems for optimization.

[0006] Various embodiments of the present invention include methods and systems for generic optimization of problems by an approach to minimizing functions over high-dimensional domains that mathematically model problems for which near optimal or optimal solutions are sought. These embodiments receive a mathematical description of a system, in symbolic form, that includes decision variables of various types, including real-number-valued, integer-valued, and Boolean-valued decision variables. The objective function with high dimensional domain that is minimized in order to find near optimal and optimal solutions may be accompanied by a variety of constraints on the values of the decision variables, including inequality and equality constraints. Various embodiments of the present invention incorporate the objective function and constraints into a global objective function. The local and global minima of this function occur at critical points. One or more of these critical points is a near optimal or optimal solution for the modeled problem. Various embodiments of the present invention transform the global objective function and a procedure for finding critical points into a system of differential equations in terms of continuous variables and parameters, so that powerful, polynomial-time methods for solving differential equations can be applied for identifying critical points of the function. These embodiments undertake an interior-point-based method and employ a global gradient-descent field formulation in order to efficiently traverse the hyperdimensional domain of the global objective function to find local minima. Near a local critical point, where the global descent field is ineffective in providing better approximants of the variables, embodiments of the present invention provides a steering method that generates a local descent-field across the local critical point towards the solution. Once the vicinity of the local critical point is left behind, the global gradient descent method is resumed, and various embodiments of the present invention employ a method to establish a reasonable direction for resuming a descent path through the hyper-dimensional domain toward local minima. One of these local minima may represent a near optimal solution of the problem. Embodiments of the present invention also provide for distribution and decomposition of the global gradient descent-field and local gradient descent-field optimization methods using multiple threads and agents, respectively, in order to allow for parallel computation and increased time efficiency. Various embodiments of the present invention further include approaches for adjusting solutions to optimization problems relatively continuously in time, without needing to recompute the optimization solution de novo. While many embodiments of the present invention are specifically directed to various classes of optimization problems, other embodiments of the present invention provide a more general approach for constructing complex hierarchical computational processes and for optimally or near optimally controlling general computational processes.

[0064] The present invention relates to the generic optimization of mathematically-formulated problems, relatively continuous adjustment of solutions to optimization control problems, and to a more abstract, higher-level hierarchical construction of, an optimal or near-optimal control of computational processes. Various embodiments of the present invention include complex software programs running on a single-processor computer system, or running, in parallel, on multi-processor computer systems, or on a large number of distributed, interconnected single-and/or-multiple processor computer systems. These programs are quite mathematical in nature, and involve elaborate and difficult-to-describe mathematical techniques. For this reason, the present invention is described in part, below, with reference to a concrete problem, with reference to many hundreds of equations, and with reference to many graphical illustrations and control-flow diagrams. Although mathematical expressions, alone, may be sufficient to fully describe and characterize the programs that implement processes that represent various embodiments of the present invention to those skilled in the art of software development, the more graphical, concrete-problem-oriented, and control-flow-diagram approaches included in the following discussion are intended to illustrate various embodiments of the present invention in a variety of different ways that may be particularly accessible to readers with particular backgrounds. The present invention is described below in the following subsections: (1) Optimization; (2) Repair; and (3) Meta-Control and Hierarchical Process Construction.

[0065] FIGS. 1A–B illustrate a very simple, two-dimensional optimization problem. FIGS. 1A–B show two different views of a two-dimensional surface, with the surface mathematically expressed as follows:
z=f(x1,x2)
where the domain of the problem is the two-dimensional surface and the goal of the problem is to identify values of x1 and x2 for which the dependent variable z is minimal. In other words, the graph of the surface is a collection of points (x1, x2, z) characterized by a function of two variables, x1 and x2, plotted with respect to the x1-axis 102, the x2-axis 104, and the z-axis 106. As a concrete example, z might represent the cost associated with manufacturing widgets of length x1 and width x2. An optimization problem with regard to this model may be to find the optimal widget length and widget width in order to reduce the cost of manufacturing a widget to the lowest possible value. In most cases, a closed-form expression for the solution, or even a mathematical model, for an optimization problem is not available. Instead, one may begin with some initial, feasible values for the decision variables, in the present case x1 and x2 values corresponding to a point on the two-dimensional surface 110, and then attempt to find a path, denoted in FIG. 1A by the dark curve 112, from the initial, feasible position 110, to a local or global minimum value of the function z_min that occurs at a point 114 on the surface with coordinates (x1_min, x2_min, and z_min).

[0066] Note that one possible approach for finding a near-optimal or optimal solution to the problem would be to compute z for every possible pair (x1, x2). However, in general, even for extremely simple problems, such an approach would be computationally prohibitive. In many cases, the domains for decision variables, such as decision variables x1 and x2, are subsets of the real numbers, or subsets of other natural domains, which may greatly decrease the potential number of decision-variable-values for which the function would need to be evaluated in order to conduct a complete, combinatory search. However, even in these cases, the combinatory search method is generally not feasible.

[0067] It should be noted that, although many two-dimensional-surface illustrations, such as FIGS. 1A–B are used in the following discussion, they are provided only for illustration purposes. In real-world problems, there may be thousands, hundreds of thousands, or millions of variables and constraints, leading to extremely high-dimensional problem domains. In general, these problem domains may be considered to be hyper-dimensional volumes, or manifolds, bounded by hyper-dimensional surfaces specified by the constraints. It is not possible to illustrate such problem domains, but techniques used to address hyper-dimensional problem domains may analogized to 3-dimensional illustrations. For example, a feasible point in a two-dimensional problem domain is a point of a two-dimensional surface, while a feasible point in a hyper-dimensional problem domain is a point within the hyper-dimensional volume, or manifold, enclosed by the hyper-dimensional boundary surface.

[0068] Many other approaches are possible for traversing the domain of a mathematical model from an initial, feasible point to a local or global minimum point. One can, for example, compute the gradient for the function at each point, and proceed along the surface in the direction of the gradient in order to most steeply descend along the surface to a local or global minimum. This approach is best suited for convex problems, in which a single global minimum exists. The hyper-dimensional volume domain for a real-word optimization problem is quite often non-convex, and includes many critical points, including saddle points, local maxima, and local minima in addition to a global minimum. Therefore, a steepest-descent approach using gradients may be used to traverse the hyper-dimensional problem domain to a local minimum, but a searching method must nonetheless be employed in order to visit and evaluate multiple local minima. For example, in FIG. 1A, had the trajectory dictated by the steepest-descent method ended at the local minimum represented by point 116, then some other approach would be needed in order to detect and evaluate the global minimum at point 112. FIG. 1B shows the problem-domain surface and path to the global minimum shown in FIG. 1A rotated by approximately forty-five degrees about the z-axis.

[0069] FIG. 2 is a control-flow diagram for a highest level view of an optimization method and system that represents one embodiment of the present invention. In a first step, 201, a particular system or problem to be optimized is mathematically modeled, and the mathematical model serves as input to subsequent optimization steps. Many of the embodiments of the present invention include graphical user interfaces, text-based interpreters, and other such interfaces in order to facilitate mathematical modeling of problems. Below, a particular mathematical model for a hypothetical problem is provided as an example of a mathematical model for an optimization problem. In general, the mathematical model is provided in symbolic form, comprising a set of mathematical expressions including symbolic terms and operators. Next, in step 202, the method and system undertake a preprocessing step in order to transform the symbolically expressed mathematical model to a standard form needed by methods that represent embodiments of the present invention. Preprocessing steps are discussed below with respect to the hypothetical mathematical model, and the preprocessing steps are formally described. Next, in step 203, the symbolic standard form of the mathematical model is used to set up an iteration and to iteratively optimize the function obtained using an instance of the interior-point method and a gradient-steered descent-field trajectory in order to identify one or more local minima within a hyper-dimensional problem domain. Various strategies and descent-field mechanisms are employed to find and evaluate a possibly large number of local minima in order to identify a near-optimal solution to the mathematically modeled problem. Various embodiments of the present invention employ different search strategies. It is important to note that neither the methods that represent the embodiments of the present invention, nor any other optimization methods currently employed, are guaranteed to always find a global minimum within a given tolerance or a reasonable amount of time using a reasonable amount of computer resources. However, most practical optimization problems are bounded, in that the number of local minima is finite. Finally, once a best solution is found within some specified tolerance or the maximum amount of time using a specified amount of computing resources, this solution is transformed, in a post-processing step 204, in order to provide one or more near-optimal solutions in the context of the mathematical model for the problem originally provided in step 201.

[0070] The mathematical modeling step and preprocess step, 201 and 202 in FIG. 2, respectively, are perhaps best appreciated using a concrete, hypothetical problem. FIG. 3 shows a hypothetical problem used to illustrate the mathematical modeling and preprocessing steps. The hypothetical problem involves p manufacturing plants, 302–304, w warehouses, 306–311, and c customers, 314–318. A manufacturing plant, such as manufacturing plant 302, can ship goods to any of the w warehouses, indicated in FIG. 3 by various types of lines, such as dashed line 320 connecting manufacturing plant 302 and warehouse 306. Each warehouse, in turn, can ship goods to any of the c customers, with paths between warehouses and customers indicated by various types of lines, such as the dash line 322 interconnecting warehouse 306 and customer 314. In FIG. 3, a different style of line is used to show the path from a particular manufacturing plant to the various warehouses, and from a particular warehouse to the various customers.

[0071] FIGS. 4–7 illustrate further aspects and characteristics of the hypothetical problem using the illustration conventions of FIG. 3. For example, the notation x_ij^pw indicates the amount of goods shipped from manufacturing plant i to warehouse j and terms of the form x_jk^wc indicate the amount of goods shipped from warehouse j to customer k. Although there are w potential warehouses, not all w warehouses need to be used. A Boolean variable b_j indicates whether or not warehouse j is employed, with the Boolean value “1” or “TRUE” indicating that the warehouse is used, and the Boolean value “0” of “FALSE” indicating that the warehouse is not used. In FIG. 4, for example, warehouses 1 (306), 3 (308), and 5 (310) are used, as indicated by the solid borders of the squares representing used warehouses and the Boolean values “1” associated with the used warehouses, and warehouses 2 (307), 4 (309), and all warehouses following warehouse 5, including warehouse w 311, are not used. The used warehouses are alternatively numbered, in FIG. 4, with sequential numbers in parentheses so that the three used warehouses are sequentially numbered 1, 2, and n, where n is the total number of used warehouses and equals 3 in the example shown in FIG. 4.

[0072] FIG. 5 illustrates costs associated with shipping goods from manufacturing plants to warehouses and from warehouses to customers. A term of the form c_ij^pw indicates the per item cost of shipping goods from a manufacturing plant i to warehouse j, and a term of the form c_jk^wc indicates the per item cost of shipping goods from warehouse j to customer k. In FIG. 5, each of the shipping routes is labeled with the corresponding per-item cost.

[0073] FIG. 6 illustrates additional characteristics of the hypothetical problem. With each manufacturing plant, there is an associated manufacturing capacity, with a term having the form s_i indicating the shipping capacity of manufacturing plant i. Each warehouse has a receiving capacity, expressed by a term v_j indicating the capacity of warehouse j to receive goods. Also, there is a cost of operation for each warehouse, with a term of the form r_j indicating the cost of operating warehouse j. Finally, there is a demand of goods associated with each customer. A term of the form d_k indicates the demand for goods by customer k. In

[0074] FIG. 6, the supply capacities, receiving capacities, cost of operations, and demands associated with manufacturing plants, warehouses, and customers are all explicitly detailed. Finally, FIG. 7 shows a simple two-manufacturing-plant, three-potential-warehouse, five-customer hypothetical problem associated with numerical costs, capacities, and demands.

[0075] The optimization problem illustrated in FIGS. 3–7 is to figure out the lowest cost configuration of operating warehouses, manufacturing-plant-to-warehouse shipping volumes, and warehouse-to-customer shipping volumes such that all the customer demands are met and no capacity is exceeded. A more formal, symbolic mathematical expression of the optimization problem follows.

[0076] First, the total cost associated with any particular system configuration is expressed as:

[0077] $\mathrm{total}\mathrm{cost}=\sum _{i=1}^p\sum _{j=1}^n{c}_{\mathrm{ij}}^{\mathrm{pw}}{x}_{\mathrm{ij}}^{\mathrm{pw}}+\sum _{j=1}^n\sum _{k=1}^c{c}_{\mathrm{jk}}^{\mathrm{wc}}{x}_{\mathrm{jk}}^{\mathrm{wc}}+\sum _{j=1}^w{r}_j{b}_j$
In other words, the total cost is the sum of the costs associated with all shipments from manufacturing plants to warehouses, from warehouses to customers, and the costs of operating warehouses. A minimum is sought in the hypothetical optimization problem for the total cost. In addition, the model includes various constraints. First, a manufacturing plant cannot ship a greater amount of goods than the capacity of the manufacturing plant, expressed as follows:

[0078] $\sum _{j=1}^n{x}_{\mathrm{ij}}^{\mathrm{pw}}\le s_i$
Customers all receive at least an amount of goods equal to their demands, expressed as follows:

[0079] $\sum _{j=1}^n{x}_{\mathrm{jk}}^{\mathrm{wc}}\ge d_k$
Furthermore, the amount of goods sent by a manufacturing plant to each warehouse is less than or equal to the warehouse's capacity for receiving goods, expressed as follows:

[0080] $\sum _{i=1}^p{x}_{\mathrm{ij}}^{\mathrm{pw}}\le v_j$
Similarly, a warehouse ships an amount of goods less than or equal to the warehouse's capacity to receive goods, expressed as follows:

[0081] $\sum _{k=1}^c{x}_{\mathrm{jk}}^{\mathrm{wc}}\le v_j$
Warehouses do not store goods. Each warehouse ships out everything that it receives, expressed by the conservation constraint that follows:

[0082] $\sum _{i=1}^p{x}_{\mathrm{ij}}^{\mathrm{pw}}-\sum _{k=1}^c{x}_{\mathrm{jk}}^{\mathrm{wc}}=0$
The shipment pathways illustrated in FIGS. 3–7 are unidirectional or, in other words, goods are not returned, expressed as follows:
x_ij^pw≧0 and x_jk^wc≧0
For the hypothetical problem it is assumed that there is sufficient manufacturing capacity and warehouse-receiving capacity to satisfy the customer demands, expressed as follows:

[0083] $\sum _{i=1}^p{s}_i\ge \sum _{j=1}^w{v}_j\ge \sum _{k=1}^c{d}_k$
Goods may be shipped to and from only operating warehouses, expressed by the following two constraints:
x_ij^pw(1−b_j)=0
x_jk^wc(1−b_j)=0
The number of operating warehouses is less than or equal to the total potential number of warehouses w:

[0084] $\sum _{j=1}^n{b}_j\le w$
Finally, the terms y_j are Boolean values, expressed as follows:
b∈{0,1}

[0085] The above mathematical model may be slightly transformed to more concisely express the cost function and associated constraints. First, the amount-of-goods-shipped parameters of the types x_ij^pw and x_jk^wc may be replaced by scaled shipment terms as follows:

[0086] $m_{\mathrm{ij}}^{\mathrm{pw}}=\frac{x_{\mathrm{ij}}^{\mathrm{pw}}}{s_i},0\le m_{\mathrm{ij}}^{\mathrm{pw}}\le 1$ $m_{\mathrm{jk}}^{\mathrm{wc}}=\frac{w_{\mathrm{jk}}^{\mathrm{wc}}}{d_k},0\le m_{\mathrm{jk}}^{\mathrm{wc}}\le 1$
The warehouse-capacity-based constraints, provided above, turn out to be redundant, and can therefore be removed from the set of constraints without affecting the problem model or solutions obtained for the model. The conservation constraint indicating the warehouses do not store goods is modified to use scaled shipment terms, as follows:

[0087] $\sum _{i=1}^p{s}_i{m}_{\mathrm{ij}}^{\mathrm{pw}}\ge \sum _{k=1}^c{d}_k{m}_{\mathrm{jk}}^{\mathrm{wc}}$
Non-operable warehouses do not ship goods, as expressed by the following constraint:
m_jk^wc≦b_j
The total cost may be expressed, using the scaled shipment of goods terms, as follows:

[0088] $\sum _{i=1}^p\sum _{j=1}^n{s}_i{c}_{\mathrm{ij}}^{\mathrm{pw}}{m}_{\mathrm{ij}}^{\mathrm{wc}}+\sum _{j=1}^n\sum _{k=1}^c{d}_j{c}_{\mathrm{jk}}^{\mathrm{wc}}{m}_{\mathrm{jk}}^{\mathrm{wc}}+\sum _{j=1}^n{r}_j{b}_j$
The capacity constraints are expressed, in terms of the scaled shipment of goods terms, as follows:

[0089] $\sum _{j=1}^n{m}_{\mathrm{ij}}^{\mathrm{pw}}\le 1$ $\sum _{j=1}^n{m}_{\mathrm{jk}}^{\mathrm{wc}}=1$
The above-listed capacity constraints can be recast as an inequality with respect to 0, as follows:

[0090] $\sum _{i=1}^p{s}_i{r}_{\mathrm{ij}}^{\mathrm{pw}}-\sum _{k=1}^c{d}_k{m}_{\mathrm{jk}}^{\mathrm{wc}}\ge 0$
The fact that the number of operating warehouses must be less than or equal to the total potential number of warehouses, that goods cannot be returned, and that the b_j terms are Booleans, are now represented as:

[0091] $\sum _{j=1}^n{b}_j\le w$ $m_{\mathrm{ij}}^{\mathrm{pw}},m_{\mathrm{jk}}^{\mathrm{wc}}\ge 0$ $b_j\in \left\{0,1\right\}$

[0092] FIG. 8 is a more detailed control-flow diagram for the preprocess routine invoked in step 202 of the high-level control-flow diagram shown in FIG. 2. Each step of the preprocess routine illustrated in FIG. 8 is further discussed below with reference to corresponding transformations of the hypothetical mathematical model described above. First, in step 801, the preprocess routine receives the initial, symbolic, mathematically expressed model for the optimization problem, including constraints. In step 803, the preprocessing routine attempts to remove as many redundant constraints as possible. An example of the removal of redundant constraints occurs in the above transformed model, with the warehouse-capacity constraints in the original form of the model recognized as redundant and removed from the transformed model. Next, in step 805, all Boolean and integer-value decision variables, such as the Boolean variables b_j, in the hypothetical example, are transformed into floating-point variables. In step 807, the resulting symbolic, mathematically expressed model is placed in a standard form. In step 809, equality constraints are replaced by relaxed inequality constraints. Each equality constraint is relaxed with three inequality constraints using an additional positively valued variable. In step 811, an additional variable z is added, along with an additional constraint to convert the criterion into an additional constraint. Finally, in step 813, the transformed optimization problem is converted into an Interior Point Method barrier function by adding together terms. Each term is an evaluation of a barrier function in terms of an expression representing a constraint multiplied by a control variable. This Interior Point Method Barrier function contains two sets of control variables u and r. As a result of the preprocessing routine, the mathematical model is symbolically expressed in a form that can be optimized by the iterative optimization technique shown as step 203 in FIG. 2.

[0093] FIG. 9 shows the initial symbolic, mathematically expressed model for the specific optimization problem illustrated in FIG. 7. The cost function, 902, is to be optimized with respect to vectors of integer variables x and Boolean variables b. The constraints include manufacturing-plant-capacity constraints 904, customer-demand constraints 906, conservation constraints 908, constraints 910 that model the fact that no flow can occur to or from a warehouse that is not selected by the optimization, the corresponding Boolean variable being zero, a constraint that only two of the three warehouses may be operational 912, a constraint specifying that there are no return shipments of goods from customers to warehouses or from warehouses to manufacturing plants 914, and the constraint that the b_j are Boolean-valued variables 916.

[0094] The third step of the preprocessing routine, step 805 in FIG. 8, involves transforming Boolean, integer, and other non-floating-point variables to floating-point variables. Each Boolean variable b_i is replaced with a floating-point value y_i in the range [0,1] and with an additional constraint:
y_i(1−y_i)=0
The additional constraint implies that the floating-point variable y_i can only have one of the two values 0 or 1. Each integer variable is replaced by an equivalent base-2 polynomial with k terms, where k is approximately equal to the base-2 logarithm of the maximum value that the integer variable can have. Each term of the base-2 polynomial comprises a power of 2 multiplied by a Boolean coefficient, and thus the base-2 polynomial is, in turn, replaced with k floating-point variables y₀, y₁, . . . y_k−1, all within the range [0,1] along with k constraints in the form:
y_i(1−y_i)=0
The conversion of decision variables of various types to floating-point numbers is summarized formally as follows:

[0095] $x_i\in R\Rightarrow x_i\mathrm{is}\mathrm{float}$ $x_i\notin R\to x_i\mathrm{is}\mathrm{transformed}\mathrm{to}\mathrm{float}\text{:}$ $\forall x_i\in \mathrm{Bool},x_i\Rightarrow y_i\in \left[0,1\right],y_i\left(1-y_i\right)=0$ $\forall x_i\in Z^{+},x_i\Rightarrow \sum _{i=0}^k{b}_i{2}^i=b_0{2}^0+b_1{2}^1+b_2{2}^2+\dots +b_{k-1}{2}^{k-1}$ $\mathrm{where}\mathrm{maxvalue}\left(x_i\right)=2^k-1$ ${\log }_2\left(\mathrm{maxvalue}\left(x_i\right)\right)\cong k$ $y_0,y_1,y_2\dots y_{k-1}\in \left[0,1\right]\mathrm{and}{x}_i=\sum _{j=0}^{k-1}{y}_j{2}^j$ $y_0\left(1-y_0\right)=0$ $y_1\left(1-y_1\right)=0$ $⋮$ $y_{k-1}\left(1-y_{k-1}\right)=0$ $x_i\in R\Rightarrow x_i\mathrm{is}\mathrm{float}$ $x_i\notin R\to x_i\mathrm{is}\mathrm{transformed}\mathrm{to}\mathrm{floats}\text{:}$ $\forall x_i\in \mathrm{Bool},x_i\Rightarrow y_i\in \left[0,1\right],y_i\left(1-y_i\right)=0$ $\forall x_i\in Z^{+},x_i\Rightarrow \sum _{i=0}^k{b}_i{2}^i=b_0{2}^0+b_1{2}^1+b_2{2}^2+\dots +b_{k-1}{2}^{k-1}$ $\mathrm{where}\mathrm{maxvalue}\left(x_i\right)=2^k-1$ ${\log }_2\left(\mathrm{maxvalue}\left(x_i\right)\right)\cong k$ $y_0,y_1,y_2\dots y_{k-1}\in \left[0,1\right]\mathrm{and}{x}_i=\sum _{j=0}^{k-1}{y}_j{2}^j$ $y_0\left(1-y_0\right)=0$ $y_1\left(1-y_1\right)=0$ $⋮$ $y_{k-1}\left(1-y_{k-1}\right)=0$
FIG. 10 shows transformation of the hypothetical mathematical model into a model employing only floating-point variables.

[0096] Next, the mathematical model is placed in standard form. The formal definition for a standard-form symbolic problem representation is provided as follows:

• • [0097] minimize C(x)
  • [0098] subject to:

[0099] ${\tilde{g}}_i\left(x\right)\ge 0\mathrm{for}i=0\mathrm{to}{M}_1-1$ ${\tilde{h}}_i\left(x\right)=0\mathrm{for}i=M_1\mathrm{to}{M}_2-1$ $\mathrm{where}$ $x=\left[\begin{array}{c}{x}_0\\ x_0\\ x_0\\ ⋮\\ x_0\end{array}\right]\mathrm{and}{x}_0\in ℛ$
In other words, a standard form model includes a cost function C(x), a number M₁ of inequality constraints of the form
{tilde over (g)}_i(x)≧0
and a number of equality constraints of the form:
{tilde over (h)}_i(x)=0
FIG. 11 shows the mathematical model for the hypothetical problem transformed to standard form. Note that transformation of the original equality and inequality constraints to constraints with 0 on the right-hand side, as needed for the standard form, is a simple matter of adding terms to both sides of the constraint equations that are not already in standard form.

[0100] The next step in the preprocessing routine (809 in FIG. 8) is to change all equality constraints to inequality constraints. In order to do so, each equality constraint is transformed to two inequality constraints, each inequality constraint involving the left-hand expression of the original equality constraint and a positive binding variable. Transformation of an equality constraint h_j(x) to two inequality constraints is formally expressed as follows:

[0101] $h_j\left(x\right)\to \left[\begin{array}{c}{h}_j\left(x\right)+s_j\ge 0\\ -h_j\left(x\right)+s_j\ge 0\end{array}\right]\to \left[\begin{array}{c}{g}_{m+j}\left(x,s_j\right)\\ g_{m+j+1}\left(x,,s_j\right)\end{array}\right]$
In addition for each j, the constraint
s_j≧0
is added to ensure that the constraints expressing h_j(x) are feasible.

[0102] Thus, following transformation of equality constraints to inequality constraints, the total number of constraints increases by a number 2k equal to the original number of equality constraints in the standard-form model of the problem. FIGS. 12A–D illustrate the effect of binding variables on transformed equality constraints. FIG. 12A shows the original inequality constraint h_j(x) with the value 0 (1202). When the equality constraint h_j(x) is transformed to three inequality constraints, as discussed above, then the range of values for h_j(x) expands about 0 to include all values between −s_j 1204 and s_j 1206. Thus, transforming an equality into three inequality constraints essentially relaxes the equality constraint from a point to an interval. Relaxing equality constraints facilitates the iterative optimization methods, to be described below, by preventing equality constraints from overconstraining, or dominating, the initial descent towards one or more local minima. Relaxing equality constraints also facilitates an interior-point descent, since equality constraints force optimization trajectories to problem-domain boundaries. However, as the optimization begins to converge, the relaxation of equality constraints needs to be removed. This is accomplished, as shown in FIGS. 12C and 12D, by decreasing the absolute value of the barrier variable sj in order to more and more tightly constrain the value of h_j(x) to a small region symmetrically disposed about 0. The degree of relaxation of an equality constraint can be expressed by a coefficient r_i for each binding variable s_j added to the criterion. Transformation of equality constraints to inequality constraints can be formally expressed as:

• • [0103] minimize C(x)
  • [0104] constrained by:

[0105] $g_i\left(x\right)\ge 0i=0,1,\dots ,m-1$ $h_j\left(x\right)=0j=m,1,\dots ,m+k-1\Downarrow \mathrm{minimize}\tilde{C}\left(x\right)=C\left(x\right)+\sum _{i=0}^{2k-1}{r}_i{s}_i$ $\mathrm{constrained}\mathrm{by}\text{:}$ $g_i\left(x\right)\ge 0i=0,1,\dots ,m+2k-1$
FIG. 13 shows the symbolic, mathematically expressed model for the hypothetical problem following transformation of equality constraints to inequality constraints. FIG. 14 graphically illustrates the effect of the binding variables r_i. In FIG. 14, the optimization method begins at initial point 1402 and descends through a problem-domain 1404 bounded by the cup-shaped surface 1406 towards a minimum 1408. Initially, the coefficients r_i for the binding variables s_i are small, providing a large total range of values, shown in FIG. 14 by the disk-shaped region 1410, about the initial point 1402. As the descent through the problem domain approaches the minimum 1408, the coefficients r_i of the binding variables se are increased, more and more tightly constraining the solution sets until, near the optimum 1408, the transformed equality constraints again become extremely tightly constrained 1412 because as the optimization method proceeds the s_i become smaller and smaller.

[0106] The next step of the preprocessing routine involves adding an envelope variable z along with an additional constraint to the mathematical model of the problem (811 in FIG. 8). The envelope variable z is defined as follows:
z≧{tilde over (C)}(x)
This leads to the following inequality constraint:
z−{tilde over (C)}(x)≧0
Following introduction of the envelope variable z, the formal statement of the problem becomes:

• • [0107] minimize z
  • [0108] with respect to x₀, x₁ . . . , x_n−1, s₀, s₁, . . . , s_k−1, z
  • [0109] constrained by:
    g_i(x) i=0 to m−1
    g_j(x,s_j) j=0 to 2k−1
    z−{tilde over (c)}(x)≧0
    FIG. 15 shows the mathematical model for the hypothetical problem following introduction of the envelope variable z.

[0110] FIGS. 16A–D illustrate the concept of barrier functions. Assume, for example, that the problem domain for an optimization problem is the wedge-shaped volume with a parabolic base 1602 shown on FIG. 16A. An initial point for an optimization 1604 has a relatively large z-value 1601, which decreases to 0 at the minimum 1608. In FIG. 16B shows an x-y plane projection 1610 of the problem domain 1602. The optimization trajectory from the initial point 1604 to the minimal point 1608 is shown as three line segments 1612–1614. In an interior point method, when traversing the problem domain, the optimization trajectory must remain within the problem domain. In other words, the current point in a descent trajectory through the problem domain to a minimum must always remain within the problem domain, interior to the boundary of the manifold in a hyper-dimensional problem domain. For example, as shown in FIG. 16B, the initial trajectory 1612 is angled towards the edge of the volume corresponding to the problem domain, and would intersect the boundary at point 1616 and, if continued in the same direction, would pass through the boundary to points exterior to the problem domain along the dashed line segment 1618. In order to keep the trajectory of the optimization within the problem domain, it is necessary to include barrier functions within a problem model that have little or no effect within the problem domain, but rise to large values as the optimization trajectory nears the boundary surface of the problem domain, in order to deflect the trajectory back from the boundary into the problem domain. For example, in FIG. 16B, as the optimization trajectory 1612 approaches point 1616, the barrier functions added to the mathematical model begin to rapidly increase in order to drive the value of the envelope variable z to high values and therefore deflect the optimization trajectory away from the inner surface of the problem domain. Following deflection, a new optimization trajectory 1613 is established which leads, through continued optimization, to a new second approach to a boundary point 1620, just prior to which the barrier constraints again force the value of the objective function z to large values in order to again deflect the optimization trajectory back into the interior of the problem domain and to a new, final trajectory 1614.

[0111] FIGS. 16C–D illustrate the operation of the barrier functions. Assuming a current optimization trajectory 1622, as shown in FIG. 16C, consider the value of the objective function z with respect to the x-coordinate, shown in FIG. 16D, for a traversal of the problem of the domain along the current trajectory 1622 shown in FIG. 16C. As can be seen in FIG. 16D, the value of z is relatively low for values of x within the interior of the problem domain along the current trajectory 1622. However, as x decreases towards the left boundary point 1624 or increases towards the right boundary point 1626, the barrier functions have the effect of rapidly increasing the value of z 1628 and 1639, respectively. In other words, for each inequality in the model, a term involving a barrier function is added to the interior point objective function in order to constrain trajectories to the interior of the problem domain. If a trajectory leaves the problem domain, or reaches the exterior of the enclosing surface of the problem domain, the optimization methods that represent some of the embodiments of the present invention could fail.

[0112] Adding terms involving barrier functions to the model is a simple way to ensure that the optimization-trajectory stays within the problem domain. A barrier function is added for each constraint, since it is the constraints that specify the surface enclosing the problem domain. A suitable barrier-function term for the optimization techniques employed in various embodiments of the present invention is the negative of a positive variable coefficient u_i, the control variable associated with the term, times the logarithm of the constraint expression g_i, or, in other words:
−u_i log(g_i)
As the value of the constraint expression g_i decreases towards 0, the value of the logarithm of the constraint log(g_i) decreases rapidly with a corresponding rapid increase in absolute value. By taking the negative of the log (g_i), a rapidly increasing value is obtained as the constraint expression g_i closely approaches 0, approaching infinity in the limit, thus preventing the constraint from ever decreasing to 0. The variable coefficients u_i can be considered to be control variables, much like the coefficients r_i of the above-described binding variables. The u_i and r_i coefficients are modified, during an optimization, in order to steer the optimization trajectory in a most efficient manner towards local minima. However, unlike the r_i coefficients, the u_i coefficients are initially large, and are decreased by one of the embodiments of this invention. as the optimization trajectory approaches a local minimum, since local minima often fall close to, or on the surface of, the problem domain. A formal expression the addition of barrier functions to the objective function is provided as follows:

[0113] $\underset{x,s,z}{\mathrm{minimize}}F\left(x,s,z,u,r\right)=z-\sum _{i=0}^{i=m+2k}{u}_i\log \left(g_i\left(x,s,z,r\right)\right)$
FIG. 17 shows the symbolic, mathematical model for the hypothetical problem following addition of barrier constraints. The addition of the barrier functions completes the preprocessing steps embodied in the preprocessing routine in a first, described embodiment.

[0114] Next, an overview of the core optimization method that represents one embodiment of the present invention is provided. FIG. 18 illustrates the use of a global gradient descent field within a problem domain in order to provide a direction for an optimization trajectory leading from an initial point to a local minimum. At each point within the problem domain, such as point 1802, the partial derivative vector of the objective function F(x,s,z,u,r) may be calculated as a vector of first partial derivatives of the objective function with respect to the decision variables x_i, the binding variables s_i, and the envelope variable z. Similarly the Hessian of the function F(x,s,z,u,r) given as the matrix of second partial derivatives of F(x,s,z,u,r) with respect to the decision variables x_i, the binding variables s_i, and the envelope variable z can be computed at every point of the domain The gradient of the objective function is defined as the vector at each point in the domain computed as the product of the inverse of the Hessian and the partial derivative of F(x,s,z,u,r). It provides a direction 1804 from the point 1802 representing a descent direction in the value of the objective function with respect to the decision variables and binding variables. Since a gradient can be calculated for each point within the problem domain, the gradient of the objective function specifies a vector field, pointing in a direction of descent and, specifying a path to a minimum within the problem domain. In general, gradients for points, such as point 1802 in FIG. 18, far from local minima have relatively large absolute values, while gradients for points close to a local minima, such as point 1806, have gradients with relatively small absolute values.

[0115] The mathematical description of the optimization methods is facilitated by a number of notational conventions. First, the objective function

• • [0116] F(x,s,z,u,r)
    can be expressed as:

[0117] $f\left(y,u,r\right)$ $\mathrm{with}\left[\begin{array}{c}x\\ s\\ z\end{array}\right]=y\left[\begin{array}{c}{u}_0\\ u_1\\ ⋮\\ u_{m+2k}\end{array}\right]=u\left[\begin{array}{c}{r}_0\\ r_1\\ ⋮\\ r_{2k-1}\end{array}\right]=r$
Thus, the objective function can be thought of as a function of a vector y of decision variables, binding variables, and the envelope variable, the vector u, a vector of barrier-function coefficients, and the vector r, a vector of binding variable coefficients. The vector y is referred to as the “state vector,” and the vectors u and r are referred to as “control vectors.” The vector y has N elements, where N=n (decision variables)+k (binding variables)+1 (envelope variable).

[0118] FIG. 19 illustrates an optimization trajectory. The trajectory 1902 is plotted with respect to the lengths of vectors r and u as horizontal axes, and the value of the objective function F as a vertical axis. The optimization trajectory begins at an initial point 1904 and descends in the value of the objective function F towards a local minimum. During the descent, the value of the barrier functions expressed by the length of vector u, decreases towards 0, while the value of the coefficients of the binding variables, as expressed as the length of vector r, increases. The trajectory is divided into segments of length Δτ, such as a first segment 1906. During optimization, the vectors y and u are continuously recalculated to drive the value of the objective function lower. In FIG. 19, the value of the binding variable coefficients, represented as vector r, is also shown to be recalculated during each Δτ interval, although in the described embodiment of the present invention, the binding variable coefficients are adjusted only towards the end of an optimization trajectory.

[0119] The discrete optimization technique may be expressed as follows:
y_k+1=y_k+φ_k(y_k, r_k, u_k)
where k=0, 1, . . . , k_convergence. In other words, a new state-vector value y_k+1 is computed, in each iteration of a computational optimization method, from a current state-vector value y_k and a discrete function φ_k that depends on y_k, r_k, and u_k, with the iteration count k ranging from 0 to a value k_convergence at which point the state vector for a local minimum of the objective function might be determined. In a first step of optimization, the state and control vectors, y, u, and r may be parameterized by a continuous iteration parameter τ, where τ ranges from 0 to an iteration horizon T. By parameterizing vectors y, U, and r by the continuous iteration variable τ, the iteration above is transformed to a first order differential equation in which the y variable is the dependent variable and u, and r are control variables. Well-known differential-equation-based mathematical technique can be used for computing efficient optimization trajectories as solutions of this differential equation. This is discussed below. The objective function is now expressed formally as:
F(y(τ), u(τ), r(τ))
where 0≦τ≦T. For example if iteration y_k+1=y_k+φ_k(y_k, r_k, u_k) is the Newton iteration, the corresponding differential equation has the following form

[0120] $\frac{dy}{d\tau }=H^{-1}\left(y\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right)G\left(y\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right)$

[0121] It should be noted that Newton's method is but one approach to construct the differential equation providing an expression for

[0122] $\frac{dy}{d\tau }.$
It was chosen for simplicity and for ease of computation. However, any of many other approaches may be used in embodiments of the present invention for approximating

[0123] $\frac{dy}{d\tau }.$

[0124] In the following discussion, the partial derivative vector G and Hessian matrix H for the objective function F are frequently used. The partial derivative vector G is defined as follows:

[0125] $\begin{array}{c}G\left(y\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right)={\left(\frac{\partial F\left(y\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right)}{\partial y\left(\tau \right)}\right)}^T\\ =\left[\begin{array}{c}\frac{\partial F\left(y\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right)}{\partial y_0\left(\tau \right)}\\ \frac{\partial F\left(y\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right)}{\partial y_1\left(\tau \right)}\\ ⋮\\ \frac{\partial F\left(y\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right)}{\partial y_{N-1}\left(\tau \right)}\end{array}\right]\end{array}$
The Hessian matrix H is a matrix of size N×N, defined as follows:

[0126] $\begin{array}{c}H\left(y\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right)={[\frac{\partial G}{\partial y}]}^T\left(y\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right)\\ =\left[\begin{array}{ccccc}\frac{\partial G_0}{\partial y_0}& \frac{\partial G_1}{\partial y_0}& \frac{\partial G_2}{\partial y_0}& \cdots & \frac{\partial G_{N-1}}{\partial y_0}\\ \frac{\partial G_0}{\partial y_1}& \frac{\partial G_1}{\partial y_1}& \frac{\partial G_2}{\partial y_1}& \cdots & \frac{\partial G_{N-1}}{\partial y_1}\\ ⋮& ⋮& ⋮& & ⋮\\ \frac{\partial G_0}{\partial y_{N-1}}& \frac{\partial G_1}{\partial y_{N-1}}& \frac{\partial G_2}{\partial y_{N-1}}& \cdots & \frac{\partial G_{N-1}}{\partial y_{N-1}}\end{array}\right]\left(y\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right)\\ ={\left[\begin{array}{ccccc}\frac{\partial G_0}{\partial y_0}& \frac{\partial G_0}{\partial y_1}& \cdots & \cdots & \frac{\partial G_0}{\partial y_{N-1}}\\ \frac{\partial G_1}{\partial y_0}& \frac{\partial G_1}{\partial y_1}& \cdots & \cdots & \frac{\partial G_1}{\partial y_{N-1}}\\ \frac{\partial G_2}{\partial y_0}& \frac{\partial G_2}{\partial y_1}& \cdots & \cdots & \frac{\partial G_2}{\partial y_{N-1}}\\ ⋮& ⋮& & & ⋮\\ \frac{\partial G_{N-1}}{\partial y_0}& \frac{\partial G_{N-1}}{\partial y_1}& \cdots & \cdots & \frac{\partial G_{N-1}}{\partial y_{N-1}}\end{array}\right]}^T\left(y\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right)\\ =\left[\begin{array}{ccc}\frac{{\partial }^{2}F\left(y\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right)}{\partial y_0^2}& \cdots & \frac{{\partial }^{2}F\left(y\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right)}{\partial y_0\partial y_{N-1}}\\ ⋮& & ⋮\\ ⋮& & ⋮\\ \frac{{\partial }^{2}F\left(y\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right)}{\partial y_{N-1}\partial y_0}& \cdots & \frac{{\partial }^{2}F\left(y\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right)}{\partial y_{N-1}\partial y_{N-1}}\end{array}\right]\end{array}$
The normalized rate of change of the state vector y with respect to the continuous iteration variable τ,

[0127] $\frac{dy}{d\tau }=-\frac{H^{-1}G}{G^T{H}^{-1}G}\left(y\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right)$
when expressed in terms of the Hessian matrix H and vector G, or as:

[0128] $\frac{dy}{d\tau }=-\frac{w\left(\tau \right)}{\rho \left(y\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right)}$ $\mathrm{where}w\left(\tau \right)=H^{-1}G\left(y\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right)\mathrm{and}$ $\rho \left(y\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right)=G^T{H}^{-1}G\left(y\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right).$

[0129] The optimization problem can be recast into a minimum time problem which is a special case of a generic, optimal control problem. In other words, by continualization of the objective function F(y,U,r), continuous methods can be used to obtain a computational approach for computing an optimization trajectory y(τ) consisting of state-variable changes that bring the value of the objective function F(y,u,r) from an initial point to a minimum. However, the problem domain is enormous, and hyper-dimensional, and so, without further computational constraints, or meta-level constraints, although a method for descending the gradient field within the problem domain has been obtained, the problem of computationally descending the gradient field is slow convergent.

[0130] An optimal control problem is formulated. It seeks to speed-up the convergence by minimizing the convergence horizon and satisfying the requirements on the control variables. In order to control minimization of the objective function F(y(τ), u(τ), r(τ)) the optimal control formulation is used to steer, using the control variables u(τ), r(τ), the optimization trajectory by selecting values for the control variables that minimize the convergence horizon. The optimal control problem for controlling the optimization trajectory is formally expressed as follows:

[0131] $\underset{u,r}{\min }𝒥\left(u\left(\tau \right),r\left(\tau \right),T\right)={\int }_o^{T}d\tau +\frac{1}{2}{G}^T\mathrm{QG}\left(y\left(T\right),u\left(T\right),r\left(T\right)\right)$
with respect to u(τ), r(τ), τ∈[0, T], where, Q is a symmetric positive definite matrix, subject to the constraints:

[0132] $\begin{array}{cc}\frac{\partial y}{\partial \tau }=-\frac{w\left(\tau \right)}{\rho \left(y\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right)}\mathrm{for}\tau \in \left[0,T\right]& \left(1\right)\end{array}$
u_i∈(0, u_maxi) ∀i  (2)
u_i(τ+Δτ)≦u_i(τ) ∀i  (3)
r_i∈[r_mini, ∞) ∀i  (4)
r_i(τ+Δτ)≧r_i(τ) ∀i  (5)
Minimization of the objective functional for the minimum convergence horizon problem, (u(τ), r(τ), T), forces F(y(τ), u(τ), r(τ)) to converge to a minimum, where G=0, in the smallest computational horizon T, where the convergence interval is the integral of the continuous iteration variable. The solution to the optimal control problem provides an optimal control policy u*(τ) and r*(τ), τ∈[0, T] that generates a state-vector-change optimization trajectory y*(τ). The first term in the integral defining the objective function (u(τ), r(τ), T) forces the optimization trajectory to an efficient as possible trajectory y* with respect to the continuous iteration variable τ. The second term within the integral forces the vector G towards 0, thereby forcing the objective function F to a minimum. The combination of these terms therefore produces an optimal control policy u*(τ) and r*(τ) to force the objective function F to a local minimum through an optimal state-vector-change trajectory y*(τ). Now the constraints of the optimal control problem (1)–(5) are discussed. The first constraint involves the normalized steepest descent for the state-vector-change trajectory, and is the primary vehicle by which the original optimization problem is incorporated into the new minimum-time control problem. The second and third constraints ensure that the control vector values u remains non-negative and less than some maximal control value u_maxi and that the u-vector values do not increase during the optimization process. The fourth and fifth constraints are analogous to the second interior constraints with respect to the binding variable coefficients r_i. The fourth and fifth constraints ensure that the binding variable coefficients r_i do not fall below a minimum value r_mini, and that the binding variable coefficients r_i do not decrease during the optimization process.

[0133] A primary result of optimal control theory, known as the Pontryagin “Minimum Principle,” provides a framework for the minimum convergence horizon problem to produce an optimal control policy u*(τ) and r*(τ)τ∈[0, T] for finding an optimal state-vector-change optimization trajectory y*(r). This approach involves the Hamiltonian function, defined as follows:

[0134] $\mathscr{H}=1-p^T\left(\tau \right)\frac{w\left(\tau \right)}{\rho \left(y\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right)}$
The Hamiltonian has a minimum value as a function of u and r at u*(τ) and r*(τ). In other words, the Hamiltonian has a minimum value for the optimal control policy. When the control policy u*(τ) is optimal, there exists an absolutely continuous function p(τ), called the costate, such that the derivatives of the state and costate functions y(τ) and p(τ) are as follows for all τ∈[0, T]:

[0135] $\frac{d{y}^{*}\left(\tau \right)}{d\tau }=\frac{\partial \mathscr{H}\left(y^{*}\left(\tau \right),u^{*}\left(\tau \right),p\left(\tau \right)\right)}{\partial p}$ $\frac{dp\left(\tau \right)}{d\tau }=-\left(\frac{\partial \mathscr{H}\left(y^{*}\left(\tau \right),u^{*}\left(\tau \right),p\left(\tau \right)\right)}{\partial y}\right)$
satisfying the boundary conditions:

[0136] $y^{*}\left(0\right)=y_0$ $p\left(T\right)=\frac{\partial G^T\left(y\left(T\right),u\left(T\right),r\left(T\right)\right)\mathrm{QG}\left(y\left(T\right),u\left(T\right),r\left(T\right)\right)}{\partial y}$ $\mathrm{for}\mathrm{all}u\left(\tau \right),\mathscr{H}\left(y^{*}\left(\tau \right),u^{*}\left(\tau \right),p\left(\tau \right)\right)\le \mathscr{H}\left(y^{*}\left(\tau \right),u\left(\tau \right),p\left(\tau \right)\right).$
The control problem solution is summarized as follows:

[0137] $\frac{d{y}^{*}}{d\tau }=-\frac{w}{\rho }$ $\frac{dp}{d\tau }=\frac{\partial }{\partial y}\left(\frac{w}{\rho }\right)p\left(\tau \right)$ $y^{*}\left(0\right)=y_0$ $p\left(T\right)=H\left(T\right)\mathrm{QG}\left(T\right)$ $p^T\frac{w\left(y^{*}\left(\tau \right),u^{*}\left(\tau \right),r^{*}\left(\tau \right)\right)}{\rho \left(y^{*}\left(\tau \right),u^{*}\left(\tau \right),r^{*}\left(\tau \right)\right)}\ge p^T\frac{w\left(y^{*}\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right)}{\rho \left(y^{*}\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right)}$

[0138] The Hamiltonian is analogous to the Hamiltonian of a mechanical system, which expresses the total energy of a mechanical system. The function y*(τ) is analogous to specifying the lowest energy positions of the components of a mechanical system in a continuous time interval, and the costate function p*(τ) is analogous to specifying the lowest energy momentum of components of a mechanical system in a continuous time interval. Using the above-derived results allows for computation, in a forward direction, along the optimization trajectory, of the partial derivative

[0139] $\frac{\partial y^{*}}{\partial \tau }$
The partial derivative

[0140] $\frac{\partial p}{\partial \tau }$
can be computed backward from τ=T from the forward-computed y*(τ). Again, with analogy to a classical mechanical problem, the momentum of the components of a mechanical system are essentially determined from the change in position of the components, over time. Thus, momentum, or costate, needs to be computed backwards from forward computed positions.

[0141] Minimization of the Hamiltonian is itself an optimization problem, and produces the optimal control policy when the Hamiltonian is minimized. This minimization problem can also be solved using Newton's approximation method, as follows:

[0142] $\frac{\partial \mathscr{H}}{\partial u}=0\mathrm{and}\frac{\partial \mathscr{H}}{\partial r}=0\left[\begin{array}{c}u\left({\tau }_{k+1}\right)\\ r\left({\tau }_{k+1}\right)\end{array}\right]=\left[\begin{array}{c}u\left({\tau }_k\right)\\ r\left({\tau }_k\right)\end{array}\right]-{\left[\begin{array}{cc}\frac{{\partial }^2\mathscr{H}}{\partial u^2}& \frac{{\partial }^2\mathscr{H}}{\partial u\partial r}\\ \frac{{\partial }^2\mathscr{H}}{\partial r\partial u}& \frac{{\partial }^2\mathscr{H}}{\partial r^2}\end{array}\right]}^{-1}\left[\begin{array}{c}\frac{\partial \mathscr{H}}{\partial u}\\ \frac{\partial \mathscr{H}}{\partial r}\end{array}\right]\left(y\left({\tau }_k\right),u\left({\tau }_k\right),r\left({\tau }_k\right),p\left({\tau }_k\right)\right)$ $\mathrm{where}$ ${\Phi }_{\kappa }\left(y\left({\tau }_k\right),u\left({\tau }_k\right),r\left({\tau }_k\right),p\left({\tau }_k\right)\right)={\left[\begin{array}{cc}\frac{{\partial }^2\mathscr{H}}{\partial u^2}& \frac{{\partial }^2\mathscr{H}}{\partial u\partial r}\\ \frac{{\partial }^2\mathscr{H}}{\partial r\partial u}& \frac{{\partial }^2\mathscr{H}}{\partial r^2}\end{array}\right]}^{-1}\left[\begin{array}{c}\frac{\partial \mathscr{H}}{\partial u}\\ \frac{\partial \mathscr{H}}{\partial r}\end{array}\right]\left(y\left({\tau }_k\right),u\left({\tau }_k\right),r\left({\tau }_k\right),p\left({\tau }_k\right)\right)$
Thus, formulation of the minimum time problem and optimal control problem provides a method for computing state vector y and control vectors u and r in a forward direction, and computing the costate vector p in a backward direction, over a range of the continuous iteration variable τ between 0 and a horizon T. This then leads to an iterative optimization method by which the objective function F is forced to a local minimum via an optimal state-vector-trajectory y*(τ) under an optimal control policy u*(τ) and r*(τ).

[0143] FIG. 20 graphically illustrates the iterative optimization method based on the above-derived state-vector-change optimization trajectory y* and optimal control policies u* and r*. The iterative optimization method can be thought of as comprising two nested loops, an outer loop controlled by the iteration variable K and the inner loop controlled by the iteration variable k. The outer nested loop represents integration of the partial differentials of the state-vector-trajectory and control policy over a continuous iteration range of 0 to T in order to find a state vector y that minimizes the objective function F. Within each such outer-loop iteration, multiple forward computations of the control policy vectors u and r and the state vector y over small intervals of the continuous iteration variable τ are carried out in steps τ_k, where k goes from 0 to a final k value, or, in other words, τ₀, τ₁, τ₂, . . . , τ_final, and then, at τ_final≧T, the results of the computations of y(τ_k) and u(τ_k) are used to back calculate the costate vectors p(τ_k) from τ=T to τ=0. During each inner loop, the computation of the y(τ_k) values are based on prior computation of u(τ_k) and r(τ_k), and the computation of both the y(T_k) and u(τ_k) are carried out using costate values p(τ_k) computed at the end of a previous iteration of the outer loop. In the first iteration, the control costate values are approximated as a constant value.

[0144] In FIG. 20, the outer loop is represented by the outer circle 2002, and the inner-loop computations are represented by step-like radial and circumferential segments, such as step 2004. The iterative optimization process begins with initial values for the state vector, control vectors, and costate vector of y₀, u₀, r₀, and p₀, respectively 2006. The outer-loop iteration variable K is set to 0 (2008). Then, a number of step-like, inner-loop iterations are carried out, beginning with iteration k=0, each successive iteration of the inner loop incrementing the inner-loop variable k by one. In the first, inner-loop iteration, control vectors U(τ_k+1) and r(τ_k+1) are computed 2004. Then, the value of the state vector at τ=τ_k+1, y_(τk+1), is computed by a number of successive computations 2010–2016. At the end of the first iteration of the inner loop, a determination is made as to whether a local minimum has been reached. If so, then the iterative optimization method terminates, indicated in FIG. 20 by the dashed line 2018. Otherwise, a next inner-loop iteration is carried out 2020–2021. Inner-loop iterations continue until τ_k≧T 2022, at which point the costate vector values p(τ) are computed in a number of steps 2024–2030 from τ=T to τ=0, and the outer-loop iteration control variable K is incremented to conclude a full outer-loop iteration and return to the beginning point 2032 for a next outer-loop iteration.

[0145] Considering FIGS. 19 and 20 together provides a graphical, conceptual view of the iterative optimization process. Each inner-loop iteration of the optimization process can be considered to correspond to a segment in the optimization trajectory shown in FIG. 19, such as segment 1906. One full outer-loop iteration, from τ=0 to τ=T can be considered to produce a large segment of the optimization trajectory, such as the optimization trajectory segment beginning at initial point 1904 and extending down to point 1908. Multiple outer-loop iterations therefore compute a number of large optimization trajectory segments that eventually drive the state vector y to a value at which the objective function F is at least a local minimum. FIG. 21 provides additional, graphical illustration of the overall iterative optimization method. As shown in FIG. 21, a first outer-loop iteration produces the optimization trajectory segment from an initial point 2102 to point 2104. A second outer-loop iteration computes the optimization trajectory segment 2106 from point 2104 to point 2108. Third outer-loop iteration produces the optimization trajectory segment 2110 from point 2108 to local optimum 2112. Note that the value of the horizon, T, may change with each outer-loop iteration. Thus, in FIG. 21, the endpoint of each optimization trajectory segment is marked T_K, with K=1, 2, and 3, indicating that the endpoint of each outer-loop iteration serves as the initial point for each subsequent outer-loop iteration. It should also be noted that the magnitude of Δτ for each inner-loop step may also vary.

[0146] FIG. 22 is a control-flow diagram that describes, in general, the iterative optimization step (203 in FIG. 2). First, initial values for y, u, r, and p are chosen, the outer-loop iteration variable K is initially set to 0, and an initial horizon T₀ is computed in step 2202. Next, in an inner loop comprising steps 2204–2208, the control vectors u and r and state vector y are propagated forward from τ=0 to τ=T_K in τ increments of τ_k. First, in step 2204, a next value for the control vectors u and r is computed. Then, in step 2205, the state vector y is propagated from τ=τ_k to τ=τ_k+Δτ_k in step 2205. In step 2206, a determination is made as to whether the optimization has converged. If so, then the iterative optimization has completed, and returns. Otherwise, a determination is made as to whether τ_k+Δτ remains less than T_K, in step 2207. If not, then τ_k+1 is assigned to τ_k+Δτ, and the inner-loop control variable k is incremented, in step 2208. However, if τ_k is greater than or equal to T_K, then the costate values are computed backwards from T_K to 0, in step 2209, and the iteration variables and a database containing the computed values of the state vector, control vectors are updated in step 2210.

[0147] In convex problems, in which there is a single critical point, the iterative optimization method described above with reference to FIGS. 19–21 normally can converge in a single thread within a reasonable amount of time even for extremely large hyper dimensional problem domains with thousands of decision variables. However, optimization problems in applications are usually not convex, and the problem domains may feature many critical points. FIG. 23 illustrates one such problem domain. In FIG. 23, the surface boundary for the problem domain 2303 is bowl-shaped with a number of prominent central protrusions, such as protrusion 2304. These protrusions may be thought of as a small mountain range with three peaks (2304, 2306 and 2308) and a deep local valley, or bowl 2310. The global optimization amount for the problem domain lie deeply below the mountain range in the depths of the outer bowl.

[0148] FIG. 24 shows the same problem domain as shown in FIG. 23, or critical points explicitly indicated. The peaks of the mountain range 2304, 2306 and 2308 represent local maxima within the problem domain, and the low point in the small valley 2310 represents a local minimum. Saddle points within the mountain range, such as saddle points 2312 and 2314, also represent critical points. FIG. 25 shows a magnified view of a critical point, such as critical point 2304 in FIG. 24. As shown in FIG. 25, the partial differentials calculated for a number of points 2502–2504 near the critical point in the x and y directions calculated for a number of points 2502–2504 near the crucial point 2506 are all small in absolute value. Therefore, the gradient calculated for the points 2202–2304 near the critical point 2506 are also close to 0. To recapitulate, the derivative of the objective function with respect to τ,

[0149] ${\left(\frac{\partial F\left(y\left(\tau \right),u\left(\tau \right),r\left(\tau \right)\right)}{\partial y}\right)}^T=G,$
gives a means for determining the optimization trajectory near a critical point. Therefore, when the gradient is 0 or near 0, as it is for points near a critical point, then the rate of the state-vector trajectory with respect to τ is also 0. Moreover, G is sensitive near the critical point, since small local variations can cause relatively large magnitude changes in G and can even switch the sign of G. However, when G is small, then Newton's method, and other similar methods used to compute state-vector trajectories, are generally not useful for finding an optimization trajectory. In other words, the gradient field generally provides the direction for near-optimal or optimal trajectories, and therefore, while the gradient field is 0, all possible trajectories appear to be equally good. It is a situation analogous to attempting to use a compass at the magnetic north or south poles. As will be discussed later, a different method for computing the optimization trajectory for the state vector needs to be employed near critical points.

[0150] There is another consideration with regard to critical points. This consideration involves search-like strategy that needs to be employed near critical points by a global optimization method in order to attempt to identify and evaluate a number of potential minima within a problem domain. FIG. 26 illustrates a first-search technique. In FIG. 26, an iteration of the outer-loop of the optimization method to horizon T_n has carried the state vector to point 2602. This point is near critical point 2306, the peak of the mountain range. The fact that the gradient is nearly 0 at this point informs the optimization method that it is arrived at the vicinity of a critical point. However, its next step depends on the type of critical point at which it has arrived. In the case of a local maximum, such as point 2306, the optimization method needs to employ some method for perturbing the optimization trajectory away from the local maximum 2306 to continue a descent 2308 towards a local minimum. However, in the neighborhood of a local minimum, such as local 2310, the optimization method needs to employ a technique to ascend out of the local minimum 2312 to another point 2314 from which the optimization method can continue to descend 2316 towards a local minimum.

[0151] FIG. 27 illustrates a second strategy for searching a problem domain. In FIG. 27, the search method has again arrived at a point 2602 near a local maximum 2306. However, upon detecting the local maximum, the optimization method spawns a number of additional, relatively independent optimization threads which continue descending in different directions 2702–2705. One of the relatively independent threads 2704 reaches a point 2708 near a saddle point 2314. At this point, this thread, 2704 forks, or spawns, an additional thread, so that each of the resulting relatively independent threads descends from the saddle point in different directions 2710 and 2712. Another of the originally spawned, relatively independent threads 2702 reaches a local minimum 2310. At this point, rather than spawning new optimization threads, the relatively independent thread 2702 terminates, reporting the local minimum back to its parent thread 2701. Parent threads can select the lowest local minimum identified by their children and return that local minimum to their parents, resulting in a recursive search of the problem domain to identify a near optimal or optimal solution to the optimization problem. In alternative embodiments, parent threads can use information about reported local minima to truncate the problem domain, in order to modify the problem domain based on the local minima, and communicate the truncation of the problem domain to all other relatively independent threads.

[0152] By either the single-threaded technique shown in FIG. 26, or the multi-threaded technique shown in FIG. 27, the optimization method needs to have a means of determining the type of critical point that is arrived at when the gradient falls to a value near 0. The critical-point beacon for illuminating types of critical points is the Hessian matrix, described above. When the gradient is near 0, then when all elements of the Hessian matrix are negative, a local maximum has been reached. When the gradient is 0 and all elements of the Hessian matrix are positive, then a local minimum has been reached. Otherwise, when the gradient is 0, and the elements of the Hessian matrix have mixed signs, then an inflection point or saddle point has been reached. The critical-point beacon properties of the non-singular Hessian matrix are formerly described as follows:

for a point y, if G|_y≈0

• • [0153] if all eigenvalues of H are negative, then y is near a local or global maximum
  • [0154] if all eigenvalues of H are positive, then y is near a local or global minimum otherwise y is near a saddle point

[0155] A single optimization thread, as discussed above with reference to FIGS. 26 and 27, comprises a collection of hierarchically organized modules that may execute concurrently in separate computational environments in order to distribute the large computational tasks involved in propagating the state vector according to the above-described optimization method. FIG. 28 illustrates the modules and modular organization of an optimization thread. The highest-level module is referred to, in this document, as a “super” 2802. The super includes a relatively large database 2804 of intermediate results. The super is responsible for computing each successive horizon T_K, for choosing an initial point, in the case of the root thread in a multi-threaded optimization method, as described above with reference to FIG. 27, and for launching and coordinating the activities of subordinate modules. The first-level subordinate modules include a u_and_r module 2806, a y module 2808, and a p module 2810. The u_and_r module 2806 is responsible for the forward computation of the control vectors u_and_r, as shown in step 2204 in FIG. 22.

[0156] The y module 2808 is responsible for propagation of the state vector through successive intervals Δτ, shown as step 2205 in FIG. 22. The p module 2810 is responsible for computing costate vector values backwards from T_k to 0 at the end of each outer-loop iteration of the iterative optimization method, shown as step 2209 in FIG. 22.

[0157] The u_and_r module 2806 distributes the control vector forward calculation amongst m agents 2812–2815. They module 2808 and the p module 2810 each distributes its respective task amongst n agents 2816–2818 and 2820–2823, respectively. Distribution of the y propagation by y module 2808 over the n y-module agents 2816–2818 is an important approach to decomposing the overall optimization computations and providing opportunities for parallel computing. The state-vector propagation computations undertaken by the y module 2808 and its respective agents 2816-2818 are the central, and most computationally expensive, tasks involved in computing the state-vector-change optimization trajectory. In an alternative embodiment of the present invention, the p module uses an efficient, interpolation method for estimating costate vector values, and thus does not need to distribute the problem among n p-module agents. Thus, the p-module computes costate without distribution. In other alternative embodiments, the p-module may distribute costate interpolation over a small number of p-module agents.

[0158] The database 2804 maintained by the super 2802 is large, and may be contained not only in memory, but in mass storage devises directly or remotely coupled to one or more computer systems over which the modules of a thread are distributed. FIG. 29 illustrates, in part, the contents of the database maintained by the super in various embodiments of the present invention. In general, the super must maintain the values of various quantities calculated by the modules at each subinterval rk of the horizon from 0 to T. Thus, in FIG. 29, each sub-column of the first column of data structures, such as sub-column 2902 of the vector G data structure 2904, represents the value of the corresponding computed entity at a particular value of τ. In FIG. 29, sub-column 2902 represents the value of the vector G at τ=0. The super stores values of the state vector 2906, the control vectors u and r 2908–2909, the vector w 2910, the scaler ρ 2911, a Morse-flags vector 2912, the costate vector p 2914, the scaler values of the objective function F 2915, the vector G 2904, the value of the Hamiltonian 2916, the current T and iteration control variables and other such local control variables 2918, and the inverses of the Hessian matrix 2920–2927.

[0159] FIGS. 30–34 illustrate, using the illustration conventions of FIG. 28, the flow of information among the hierarchically organized modules of an optimization thread during one outer-loop iteration of the optimization method. In the case of a root thread, as shown in FIG. 30, an initial point for the optimization trajectory is determined 3002. In addition, initial values for the state vector, costate vector, control vector, and control variables, among other data structures, are determined for initializing the various data structures and elements 3004. Non-root threads inherit a starting point and current values for the various computational entities, such as the state and control vectors, from parent threads. As shown in FIG. 31, in the first step of the inner loop (2204 in FIG. 22), the u_and_r module 2806 receives the values of the state vector, costate vector, vector w, and the inverse of the Hessian from the super 3102 and distributes these values and/or portions of these values to u_and_r-module agents 3104–3107, receives intermediate computed results back from the u_and_r-module agents 3108, 3109, 3110, and 3111, computes a new control vector value, and returns the new control vector value 3112 back to the super. FIG. 32 illustrates data flow during the next step of the inner loop (2205 in FIG. 22). In this step, they module 2808 receives the newly computed control vectors u and r from the super 3202, distributes these values, along with additional values obtained from the super and/or computed by the y module, including the Hessian, vector other such values, to the y agents 3204–3206, receives intermediate computed values back from the y agents 3208 3209 3210, and computes and returns to the super computed values for the state vector, w vector, differential of the w vector with respect to the state vector, and the inverse of the Hessian, among others 3210. The y module carries out a number of individual propagation steps for each inner-loop iteration, as discussed above. FIG. 33 indicates that the data-exchange steps illustrated in FIGS. 31 and 32 are successively repeated in a number of inner-loop iterations. FIG. 34 illustrates the data exchange involved in costate vector calculations undertaken by the p module 2810 in step 2209 of FIG. 22. The p module receives computed values of the state vector, control vectors, and other such values from the super 3402, distributes these values, or portions of the values, among the p agents 3404-3407, receives intermediate results from the p agents 3408–3411, and returns computed values for the costate vector over a range of τ values from 0 to T_K 3412 to the super. As discussed above, in currently preferred embodiments, the p module interpolates, rather than integrates, costate vectors, and thus does not need to distribute the problem in the manner illustrated in FIG. 34.

[0160] In the following discussions, for notational convenience, the u and r vectors are considered together as the u vector. In other words, in the subsequent discussion, the binding variable coefficients and barrier-function coefficients together compose a single control vector u. Recall that the iterative optimization strategy employed by embodiments of the present invention involves integrating

[0161] $\frac{dy}{d\tau }\left(\tau \right),$
expressed as follows:

[0162] $\frac{dy}{d\tau }\left(\tau \right)=-\frac{w\left(y\left(\tau \right),u\left(\tau \right)\right)}{\rho \left(y\left(\tau \right),u\left(\tau \right)\right)}$ $\mathrm{where}$ $w\left(y\left(\tau \right),u\left(\tau \right)\right)={\left(H\left(y\left(\tau \right),u\left(\tau \right)\right)\right)}^{-1}G\left(y\left(\tau \right),u\left(\tau \right)\right)$ $\rho \left(y\left(\tau \right),u\left(\tau \right)\right)={\left(G\left(y\left(\tau \right),u\left(\tau \right)\right)\right)}^{T}w\left(y\left(\tau \right),u\left(\tau \right)\right)$
In one embodiment, the integration of

[0163] $\frac{dy}{d\tau }\left(\tau \right)$
is carried out by a forward Euler integration method with variable step size Δτ_k:

[0164] $y^i\left({\tau }_k+{\mathrm{\Delta \tau }}_k\right)=y^i\left({\tau }_k\right)+\Delta {\tau }_k\frac{d{y}^i}{d\tau }\left({\tau }_k\right)$
Therefore, as is apparent in the above expressions, they module, for each forward integration step, needs to invert the Hessian matrix in order to compute the w vector as the product of the inverse of the Hessian and the vector G FIG. 35 graphically illustrates this computation. Inversion of the Hessian matrix is by far the most computationally intensive task involved in optimization. Therefore, the y module distributes inversion of the Hessian matrix among the y-module agents, relying on several assumptions for this distribution.

[0165] Recall that:

[0166] $H\left(y\left({\tau }_k\right),u\left({\tau }_k\right),w\left({\tau }_k\right),u\left({\tau }_k\right)\right)={\left[\frac{\partial G}{\partial y}\right]}^T\left(y\left({\tau }_k\right),u\left({\tau }_k\right)\right)$
FIG. 36 illustrates multiplication of the Hessian matrix by the w vector to produce the g vector in a distributed fashion. As shown in FIG. 36, the G and w vectors, and the Hessian matrix, can be vertically partitioned into partitions i=0 to 5. The first partition of the G vector 3602 is referred to as G⁰, the first vertical partition of the Hessian matrix 3604 is referred to as H⁰, and the first partition of the w vector 3606 is referred to as w⁰. The Hessian matrix can additionally be partitioned horizontally into horizontal partitions j=0 to 5. In FIG. 36, the first horizontal partition of the Hessian matrix 3608 appears as a first vertical column within the Hessian matrix.

[0167] Both vertical and horizontal partitioning of the Hessian matrix results in a block-like partitioning of the Hessian matrix into submatrices. For example, FIG. 36 shows the Hessian matrix partitioned into thirty-six submatrices, beginning with the left-most and upper-most submatrix H^0,0 3610 and ending with the right-most and lower-most submatrix H^5,5 3612. Note that, in the case of a diagonally dominant Hessian matrix, in which the diagonal block submatrices H^0,0, H^1,1, H^2,2, H^3,3, H^4,4, and H^5,5 have elements with magnitude much greater than the elements of other non-diagonal submatrices, then matrix multiplication of the Hessian matrix by the w vector can be approximated by matrix multiplication of w-vector partitions by the block-diagonal submatrix within the corresponding vertical partition of the Hessian matrix. For example, in a full matrix multiplication, the G-vector partition G⁰ equals the sum of the matrix-multiplication products of the partitions of the w vector with corresponding blocks of the first vertical partition of the Hessian matrix. Formally,

[0168] $\begin{array}{c}{G}^i=H^{i}w\\ =\sum _{j=0}^{J-1}{H}^{\mathrm{ij}}{w}^j\\ H^i=\left[\begin{array}{ccc}{H}^{i0}& \cdots & H^{i\left(J-1\right)}\end{array}\right]\end{array}$

[0169] When the Hessian matrix is diagonally dominant, then the off-diagonal submatrices have relatively less magnitude and are of less importance in computation of a w-vector partition. This situation occurs when the Hessian matrix is partitioned in a way that removes, as much as possible, dependence between each partition and all other partitions. Therefore, a particular y-module agent can approximately calculate the current value of a w-vector partition 3810 by inverting a corresponding Hessian block-diagonal submatrix 3808 and multiplying the large, bracketed term 3812 in FIG. 38 by the inverse of the block-diagonal Hessian submatrix 3808, using previously calculated values for the remaining w-vector partitions 3814–3817.

[0170] FIG. 39 illustrates computation of a w-vector partition w¹ by the second y-module agent. The second y-module agent 3902 obtains the current Hessian matrix block-diagonal submatrix H^1,1 and current G-vector partition G¹ and most recently calculated values of the w-vector partitions w⁰, w², w³, w⁴, and w⁵ from the y-module 2808, computes the inverse of the current Hessian block-diagonal submatrix H^1,1 and then, using the computed inverse of the Hessian matrix, computes the current approximate value of the w-vector partition w¹. The y-module agent 3902 can then, in a next computational step, fumish the approximate current value of the w-vector partition to the y-module 2808 for distribution to the other y-module agents, and to receive their most recent computed values for the other w-vector partitions w⁰, w², w³, w⁴, and w⁵ in order to compute a next w-vector partition value w¹ at the next τ increment. Using this method, the y-module agents can synchronously compute the approximate value of w-vector partitions at each step.

[0171] Next, a control-flow diagram illustration of the super module is provided. FIG. 40 is a control-flow diagram illustrating a highest conceptual level of the super module. In step 4002, the super module determines whether or not it is a replicate, or child thread, spawned or forked from a parent thread. If not, and the current thread is therefore the root thread, then in step 4004, the super module determines an initial point for the optimization and, in step 4006, initializes various computational entities, including the state vector, costate vector, and other sets of values, and also initializes a number of variables, including the Boolean variables “critical” and “precritical” to the Boolean value FALSE. Following initialization by a root thread, or as a result of flow of control from step 4002 for a replicate thread, the super module next chooses, in step 4008, a horizon T. Then, the super module invokes the function “integrate_u_and_y_to_T,” in step 4010, which carries out successive inner-loop interactions of the optimization method, described above. Next, in step 4012, the super module invokes the routine “evaluate” in order to make decisions based on the inner-loop iterations carried out by the routine “integrate_u_and_j_to_T” in step 4010. The routine “evaluate” returns a Boolean value “continue” which the super module evaluates, in step 4014, in order to determine whether or not to carry out yet another iteration of the outer loop of the optimization method, described above. If so, then the super module calls the routine “integrate p” 4016 to back calculate costate-vector values and then returns to step 4008 to initiate another outer-loop iteration. If the super module decides not to continue outer-loop durations, then the super module invokes the routing “report”, in step 4018 and then returns.

[0172] Control-flow diagrams illustrating steps 4004, 4008, 4010, 4012, and 4018 are next provided. The routine “integrate p” called in step 4016, and the routines “integrate u” and “integrate y,” called from the routine “integrate_u_and_to_T” are subsequently described mathematically, rather than using control-flow diagrams, in the interest of clarity and brevity.

[0173] FIG. 41 is a control-flow diagram for the routine “initial point” called in step 4004 of FIG. 40. In order to determine an initial value for the state vector, y₀, the initial point routine, in step 4102, guesses at a value using any supplied hints or heuristics from users or from the mathematical modeling or preprocessing portions of the optimizer. Then, in the for-loop comprising steps 4104–4107, the initial point routine evaluates each constraint g_i in order to determine whether or not the constraint, evaluated at the initially guessed state-vector point y₀, is greater than 0. If so, then the constraint is placed into partition I in step 4107. If not, then the constraint is placed in partition J in step 4106. Then, in step 4110, the initial point routine constructs an objective function {tilde over (F)} as follows:

[0174] $\tilde{F}=-\sum _{i\in I}{u}_i\log \left(g_i\left(y\right)\right)-\sum _{i\in J}{u}_j\log \left(g_j\left(y\right)+T_j\right)-\sum _{l\in J}{u}_l\log \left(T_l\right)$
which is then minimized in step 4112 by a recursive call to the optimizer (FIG. 30) to produce a modified initial state-vector value. The objective function {tilde over (F)} can be thought of as initially expanding the problem-domain boundary of the objective function F in order to envelope the initial value for the state vector, and the optimization of {tilde over (F)} then contracts the problem-domain boundary, forcing the initial point into the interior of the problem domain of the objective function F. In step 4114, the initial point routine determines whether the modified state-vector value is feasible. A feasible state-vector value is an interior point of the problem domain with a reasonably high probability of being close enough to a local minimum for the optimization method to find a local minimum. If the modified state-vector value is feasible, then it is returned as the initial point for the optimization. Otherwise, control flows back to step 4102 for another pass at determining an initial state-vector value. In subsequent iterations, information gleaned from previous iterations may be employed in step 4102 in order to better guess an initial value for the state vector.

[0175] FIG. 42 is a control-flow diagram for the routine “choose_next_T.” In step 4201, the routine “choose_next_T” assigns local variable “T” to a minimum horizon constant. Then, in the for-loop, comprising steps 4203–4207, the routine “choose_next_T” calculates, for each y-module agent, a T value obtained by dividing a constant parameter by the minimum absolute eigenvalue of the block-diagonal Hessian submatrix associated with the y-module agent. If the computed T value for the y-module agent is greater than the contents of the variable “T,” as detected in step 4205, then the variable “T” is assigned to the T value calculated for they-module agent. Next, in step 4208, the routine “choose_next_T” initializes local variables “n” to a maximum number of increments, initializes local variable “a” to an increment coefficient, and initializes a loop-control variable “j” to 0. Then, in steps 4210–4212, the routine “choose_next_T” increases the value of the variable “T” in step 4212 until the derivative of the costate vector with respect to T evaluated for the current contents of the variable “T” approaches 0, as detected in step 4210. The loop may also terminate when the number of iterations is equal to the contents of the local variable “n.” When

[0176] $\frac{dp}{dT}=0,$
then computation of the costate vectors over the continuous iteration range 0 to T can be more easily calculated. In alternative embodiments of the present invention, the costate vectors can be computed using a polynomial approximation, rather than employing complex integration.

[0177] FIG. 43 is a control-flow diagram of the routine “evaluate,” called in step 4012 of FIG. 40. In step 4302, the routine of “evaluate” sets the Boolean flags “at_minimum” and “continue” to the Boolean value FALSE. Next, in step 4304, the routine evaluate determines whether the computational resources allocated for the optimization have been exhausted. This may include continuous computation for longer than a maximum specified time, the usage of processing resources, memory, or a combination of computational resources greater than a specified maximum amount of resources to be used, the occurrence of error conditions, or other types of predetermined termination criteria. If the computational resources are exhausted, then the routine evaluate returns. Otherwise, the routine evaluate, in step 4306, determines whether the optimization thread is currently critical. An optimization thread is critical if all y-module agents are currently in Morse mode, and have been so for some predetermined number of iterations. Morse mode is described in detail, below. If the current thread is not critical, then the routine “evaluate” sets the Boolean flag “continue” to TRUE, in step 4308, and then returns. Otherwise, in step 4310, the routine “evaluate” determines whether the thread, currently determined to be critical, has reached a local minimum. If so, then the Boolean flag “at-minimum” is set to TRUE, in step 4312, and the routine “evaluate” returns. If the optimization thread is not at a local minimum, but is critical, then the routine “evaluate,” in step 4314, determines a number of replicant threads to spawn, and also initial directions or starting points for the replicant threads. Spawning of replicant threads is illustrated in FIG. 27. In the for-loop of steps 4316–4319, the determined number of replicant threads are spawned, each replicant thread linked has a child to the current optimization thread. Then, in step 4320, the routine “evaluate” may attempt to perturb the current optimization trajectory in order to move the optimization trajectory away from the critical point and continue descending towards a local minimum. In step 4322, the routine “evaluate” sets the Boolean flag “continue” to the Boolean value TRUE that then returns.

[0178] FIG. 44 is a control-flow diagram for the routine “report.” This routine is invoked in step 4018 of FIG. 40. In step 4402, the routine “report” determines whether the current optimization thread has a parent. If not, and the current optimization thread is therefore the root optimization thread, then the routine “report” determines, in step 4404, whether there are any children threads still running that are linked to the current, root optimization thread. If so, then the current, root optimization thread waits, in step 4406, for the children optimization threads to finish execution. Next, in step 4408, the routine “report” sets local variable “min” to the current value of the objective function computed with the current state-vector value, sets the local variable “location” to the current state-vector value, and sets the local variable “at min” to the Boolean value flag “at_minimum” returned by the routine “evaluate” for the current optimization thread. In step 4410, the routine “report” determines whether any children threads have terminated and reported back to the current optimization thread. If so, then in the for-loop of steps 4412-4415, the current optimization thread evaluates the report generated by each terminated child in order to determine the lowest value for the objective function reported by any child and achieved by the currently executing optimization thread. Next, in step 4416, the routine “report” determines whether there are any children still running. If so, then in thefor-loop of steps 4418–4420, the routine “report” links each child to the parent of the currently executing thread. Note that the test in step 4416 fails in the case that the currently executing optimization thread is the root thread. If the currently executing optimization thread is the root thread, as determined in step 4422, then the routine “report” returns the minimum objective function value, state vector, and indication of whether the optimization is currently at a minimum in step 4424. Otherwise, the routine “report” reports these same values back to its parent thread, in step 4426.

[0179] FIG. 45 is a control-flow diagram for the routine “integrate_u_and_to_T,” called in step 4010 of FIG. 40. In the for-loop of steps 4502–4507, the routine “integrate_u_and_y_to_T” successively calls the routines “integrate u” and 'steps 4503 and 4504, in order to complete one inner-loop execution for the optimization method. The successive iterations continue until either the optimization thread goes critical, as detected in step 4505, until the value τ_k is greater than or equal to T, or until some other stopping criteria are detected, such as exhaustion of computational resources or other such stopping criteria in step 4507.

[0180] FIG. 46 is a control-flow diagram for the routine “integrate y.” This routine describes the operation of the y module (2808 in FIG. 28). In step 4602, the routine “integrate y” obtains symbolic representations of the Hessian matrix, the vector G, and other data needed by the routine “integrate y” from the super module. Next, in step 4604, the routine “integrate y” computes numerical values for the Hessian matrix and vector G. In step 4606, the routine “integrate y” distributes partitions of the computed Hessian and G vector to each y-module agent. The y-module agents compute their associated w-vector partition values and distributed row values, which are collected by the routine “integrate y” in step 4608. In step 4610, the routine “integrate y” determines whether any agents are currently in Morse mode.

[0181] The term “Morse mode” refers to an agent that currently needs to compute its w-vector partition based on the Morse quadratic approximation of the original objective function associated with the y agent method since the agent is currently near a critical point with respect to the agent's partition of the vector G. If any agents are in Morse mode, then, in step 4612, the routine “integrate y” directs the Morse-mode agents i to recompute pⁱ and wⁱ and collects the recomputed pⁱ and wⁱ from each Morse-mode agent i. Next, in step 4614, the routine “integrate y” determines whether all agents are currently in Morse mode. If so, then, in step 4616, the routine “integrate y” determines whether the local variable “precritical” currently has the Boolean value TRUE. If so, then the agents were previously in Morse mode in the previously executed iteration, and so the local variable “critical_count” is incremented, in step 4618. In step 4619, the routine “integrate y” compares the contents of the local variable “critical_count” to a threshold constant and, when the contents of the local variable “critical_count” equals or exceeds the critical constant, the routine “integrate y” concludes that the current optimization thread has remained critical for a sufficient number of iterations to declare the optimization thread to be critical by setting the Boolean flag “critical” to Boolean value TRUE in step 4620. However, if the contents of the variable “critical_count” do not justify declaring the current optimization thread to be critical, then control flows to step 4624. If, back in step 4616, the routine “integrate y” determines that the variable “precritical” did not have the Boolean value TRUE, then, in step 4622, the variable “precritical” is set to true and the variable “critical_count” is initialized to 0 before control flows to step 4624. In step 4624, the routine “integrate y” computes global ρ and w-vector values, and it computes from them the value of

[0182] $\frac{dy}{d\tau }.$

[0183] Next, in step 4625, the routine “integrate y” determines a value Δτ for the current Euler forward integration step. If the computation of Δτ succeeds, as determined in step 4626, then the routine “integrate y” carries out the forward integration step, in step 4628, updates the binding variable coefficients, if necessary, in step 4630, and returns. If the computation of Δτ fails, as detected in step 4626, then the routine “integrate y” returns with an error.

[0184] The forward Euler integration technique has been discusses above, but is presented formally below. The global value for ρ is assembled from the partitioned ρ values computed by y-module agents in a current inner-loop step as follows:

[0185] $\rho \left(y\left({\tau }_k\right),u\left({\tau }_k\right)\right)=\sum _{i=0}^{n-1}{\rho }^i\left(y\left({\tau }_k\right),u\left({\tau }_k\right)\right)$
Next, the values for the partitions of the w vector are collected from the y-module agents and assembled to produce a w vector. Then, a continuous iteration increment Δτ is computed, by a method to be described below. Finally, the state vector is propagated from a current continuous iteration value τ_k to a next continuous iteration value τ_k+Δτ as follows:
yⁱ(τ_k+Δτ_k)=yⁱ(τ_k)+Δ_kdyⁱl /dτ(τ_k)
Finally, the continuous iteration value τ_k is updated as follows:
τ_k=τ_k+Δτ_k
Note that ay-module agent is considered to be critical with respect to the y-module's vector-G partition when the absolute value of the y-module agent's local ρ, ρⁱ, falls below some constant parameter. The computation of a w-vector partition wⁱ and the y-module-agent-local ρ, ρⁱ, is described below.

[0186] FIG. 47 is a control-flow diagram for the routine “compute Δτ,” called in step 4625 of FIG. 46. In step 4702, a guess, or estimate, Δτ′, is made for Δτ. This guess or estimate may be obtained from a constant parameter, or may be, in more elaborate embodiments, derived by empirical methods or heuristics from prior computed values of Δτ and from recent history of state-vector values, control-vector values, and other such information. In the described embodiment, for forward Euler integration steps, when the control vector has an approximate constant value, then F decreases by one at each step:

[0187] $\frac{dF}{d\tau }\left(y\left({\tau }_k\right),u\left({\tau }_k\right)\right)=-1.$
When some of the y-module agents are in Morse mode,

[0188] $\frac{dF}{d\tau }$
may depart from the value “−1.” However, if the value of

[0189] $\frac{dF}{d\tau }$
ever rises above 0, as detected in step 4704 of the routine “compute Δτ,” then an error condition obtains since

[0190] $\frac{dy}{d\tau }\left({\tau }_k\right)$
is not a descent direction and therefore the forward Euler integration method cannot produce a feasible trajectory y*(τ). Otherwise, the estimate of Δτ, Δτ′, is inserted into the forward Euler integration step equation

[0191] $y\left({\tau }_k+\Delta {\tau }^{\prime }\right)=y\left({\tau }_k\right)+\Delta {\tau }^{\prime }\frac{dy}{d\tau }\left({\tau }_k\right)$
and the state vector is propagated to y(τ_k+ Δτ′). This computed next state-vector value is checked, in step 4706, for feasibility. If the state-vector value is not feasible, then the estimate of Δτ, Δτ′, is decreased, in step 4708, and a new forward state-vector value is again computed and checked for feasibility in step 4706. Once a feasible forward state-vector is found with a corresponding Δτ estimate, Δτ′, then, in step 4710, the Armijo condition

[0192] $F\left(y\left({\tau }_k+\Delta {\tau }^{\prime }\right),u\left({\tau }_k\right)\right)-F\left(y\left({\tau }_k\right),u\left({\tau }_k\right)\right)<\sigma \Delta {\tau }^{\prime }{G}^T\left(y\left({\tau }_k\right),u\left({\tau }_k\right)\right)\frac{dy}{d\tau }\left({\tau }_k\right)$
is evaluated with respect to the estimate of Δτ, Δτ′. If the Armnijo condition is satisfied, then the estimate of Δτ, Δτ′, is returned as the new computed Δτ. Otherwise, Δτ is approximated. Several different approximation methods are available. In the case that a previous feasible estimate of Δτ, Δτ^A, is available, then a quadratic approximation to the objective function F around τ_k can be made, as follows:
F(y(τ_k+Δτ), u(τ_k))≈{tilde over (F)}(Δτ)=aΔτ²+bΔτ+c
Then, an estimate for Δτ is obtained as follows:

[0193] $\tilde{F}\left(0\right)=F\left(y\left({\tau }_k\right),u\left({\tau }_k\right)\right)=c$ $\tilde{F}\left(\Delta {\tau }^A\right)=F\left(y\left({\tau }_k+\Delta {\tau }^A\right),u\left({\tau }_k\right)\right)={a\left(\Delta {\tau }^A\right)}^2+b\left(\Delta {\tau }^A\right)+c$ $\frac{d\tilde{F}}{d\Delta \tau }\left(0\right)={\left(G\left(y\left({\tau }_k\right),u\left({\tau }_k\right)\right)\right)}^T\frac{dy}{d\tau }\left({\tau }_k\right)=b$ $c=\tilde{F}\left(0\right)$ $b={\left(G\left(y\left({\tau }_k\right),u\left({\tau }_k\right)\right)\right)}^T\frac{dy}{d\tau }\left({\tau }_k\right)$ $a=\frac{\tilde{F}\left(\Delta {\tau }^A\right)-\tilde{F}\left(0\right)-\Delta {{\tau }^A\left(G\left(y\left({\tau }_k\right),u\left({\tau }_k\right)\right)\right)}^T\frac{dy}{d\tau }\left({\tau }_k\right)}{{\left(\Delta {\tau }^A\right)}^2}$
A Δτ along the descent direction that minimizes F can be approximated by finding a Δτ that minimizes {tilde over (F)}, which occurs when: