Simulation system using model

[0002] This continuing application is filed under 35 U.S.C. §111(a), based upon International Application PCT/JP98/04750, filed Oct. 20, 1998, which PCT/JP98/04750 application in turn claims priority based upon Japanese Patent Application 10/104129, filed Mar. 31, 1998.

[0003] 1. Field of the Invention

[0004] The present invention relates to a simulation apparatus and the like and, more particularly, to a simulation apparatus which can efficiently model and simulate a system including complicated combinations of a plurality of elements or components like a power train of a vehicle, and the like.

[0005] 2. Description of the Related Art

[0006] For the purpose of shortening the time required for product development, it is becoming increasingly important to perform estimation and confirmation of performance and functions, which are conventionally done by tests using samples, during the initial period of development.

[0007] Various kinds of performance of a vehicle appear as integrated characteristics of individual units such as an engine, gear box, and the like. Since these characteristics stretch across a plurality of different fields, different theoretical systems are combined and modeled by a unique method that meets the purpose of simulation. However, it is not efficient to fulfill simulations against many problems and to maintain and manage them.

[0008] To solve this problem, a modeling technique which is common to different fields and can easily integrate them is required.

[0009] The present inventors have proposed a technique for expressing each component by concepts called potential and flow variables using a system equation in Japanese Patent Application No. 7-250974 and the like, thereby expressing a unit built by a plurality of components as a matrix.

[0010] The present invention has been made as an extension of and based on the above proposal, and has as its object to provide a simulation apparatus which can simulate while easily coping with changes in component and model.

[0011] In order to achieve the above object, as proposed in claim 1, a simulation apparatus for simulating, using a system model, dynamic performance of a target apparatus to be simulated including a driving source and a plurality of units for transmitting a driving force output from the driving source, said simulation apparatus comprising:

[0012] functional modeling means for representing each of the units included in the target apparatus by exchange of energy by dynamic performance of an individual unit and a parameter that determines the dynamic performance of the unit;

[0013] data storage means for storing static characteristic data of each unit; and

[0014] data modeling means for computing a parameter of each unit on the basis of static characteristic data for the unit read out from said data storage means.

[0015] According to claim 2 as a preferred aspect of the present invention, the parameter of each unit represents at least one of a mass, spring, and damper. In an embodiment to be described later, the mass is an inertial mass, the spring is rigidity, and the damper is a viscous coefficient.

[0016] According to claim 3 as a preferred aspect of the present invention, the functional modeling means represents the dynamic performance of each unit by a matrix.

[0017] According to claim 4 as a preferred aspect of the present invention, the apparatus further comprises further comprising means for integrating matrices representing dynamic performances of the individual units generated by said functional modeling means into a matrix representing the entire target apparatus.

[0018] According to claim 5 as a preferred aspect of the present invention, the dynamic performance of each unit is expressed by a state equation.

[0019] According to claim 6 as a preferred aspect of the present invention, the data modeling means converts static characteristic data read out from the data storage means into a parameter in accordance with a progressive state of a simulation of the unit.

[0020] According to claim 7 as a preferred aspect of the present invention, the individual units are represented as being coupled to each other via a potential variable that expresses a potential, and a flow variable that expresses a flow.

[0021] According to claim 8 as a preferred aspect of the present invention, the parameter includes a state variable of each unit, and a state change variable upon infinitesimal state transition. Since the state change is that in the infinitesimal state, even when each unit is nonlinear, accurate simulation can be done while assuring linearity.

[0022] According to claim 9 as a preferred aspect of the present invention, each unit performs a nonlinear operation.

[0023] According to claim 10 as a preferred aspect of the present invention, the apparatus further comprises a number of data modeling means equivalent to said data modeling means, said number being set in correspondence with types of units.

[0024] According to claim 11 as a preferred aspect of the present invention, the dynamic performance is expressed by a time in a discrete system, and the parameter of each unit is approximated to a zero-th order (constant) per unit time.

[0025] The object of the present invention is also achieved by a simulation apparatus of claim 12.

[0026] Especially, according to the simulation apparatus of claim 13, since data modeling is done for each unit, simulation precision and the degree of freedom in a control system can be improved.

[0027] The object of the present invention is also achieved by a simulation method of claim 14.

[0028] FIG. 1 is a diagram for explaining a scheme for expressing a unit by potential and flow variables, which scheme has been proposed by the present inventors in Japanese Patent Application No. 7-250974, and is used in a simulation apparatus according to an embodiment of the present invention;

[0029] FIG. 2 is a diagram for explaining another scheme for expressing a unit by potential and flow variables, which scheme has been proposed by the present inventors in Japanese Patent Application No. 7-250974, and is used in a simulation apparatus according to the embodiment of the present invention;

[0030] FIG. 3 is a diagram for explaining a scheme for expressing a unit by a closed loop system;

[0031] FIG. 4 is a diagram for explaining a scheme for expressing a unit by an open loop system;

[0032] FIG. 5 is a view for explaining the concept of a virtual prototype used in the embodiment of the present invention;

[0033] FIG. 6 is a view for explaining the arrangement of a power train to which the concept of the virtual prototype shown in FIG. 5 is applied;

[0034] FIG. 7 is a view showing an engine model to which the virtual prototype is applied;

[0035] FIG. 8 is a block diagram showing a functional model of the engine shown in FIG. 7;

[0036] FIG. 9 is a view for explaining symbols used in the embodiment of the present invention;

[0037] FIG. 10 is a block diagram showing a model of a torque converter to which the virtual concept is applied;

[0038] FIG. 11 is a diagram for explaining a functional model of the torque converter shown in FIG. 10;

[0039] FIG. 12 is a view for explaining integration of system equations of a plurality of units;

[0040] FIG. 13 is a view for explaining integration of system equations of a plurality of units;

[0041] FIG. 14 is a view for explaining integration of system equations of a plurality of units;

[0042] FIG. 15 is a view for explaining deletion of connected state variables upon integration;

[0043] FIG. 16 shows a continuous-time system model of system equations of the integrated units;

[0044] FIG. 17 is a view for explaining the principle of converting the system equations of the integrated units into a discrete-time system;

[0045] FIG. 18 is a view for explaining the principle of converting the system equations of the integrated units into a discrete-time system;

[0046] FIG. 19 is a diagram showing a model of a clutch to which the virtual concept is applied;

[0047] FIG. 20 is a diagram showing a model of a clutch/brake to which the virtual concept is applied;

[0048] FIG. 21 is a sectional view for explaining the arrangement of an automatic gear box to be simulated;

[0049] FIG. 23 is a schematic view for explaining the arrangement and torque transmission in the automatic gear box shown in FIG. 21;

[0050] FIG. 24 is a block diagram for explaining the arrangement and torque transmission in the automatic gear box shown in FIG. 21;

[0051] FIG. 25 is a view showing the model concept of a turbine shaft in the automatic gear box shown in FIG. 21;

[0052] FIG. 26 is a block diagram showing a functional model of the turbine shaft shown in FIG. 25;

[0053] FIG. 27 is a graph showing the response characteristics of the turbine shaft having the function shown in FIG. 26;

[0054] FIG. 28 is a view for explaining a functional model of a clutch to which the virtual concept is applied;

[0055] FIG. 29 is a graph for explaining the characteristics of the clutch model shown in FIG. 28;

[0056] FIG. 30 is a view for explaining gear rotations in a ravigneaux planetary gear unit used in the embodiment of the present invention;

[0057] FIG. 31 is a view for explaining torque transmission in the planetary gear unit used in the embodiment of the present invention;

[0058] FIG. 32 is a diagram for explaining a functional model of the planetary gear unit;

[0059] FIG. 33 is a perspective view for explaining the arrangement of a CR—CR planetary gear unit;

[0060] FIG. 34 is a view for explaining the torque flow in the gear unit shown in FIG. 33;

[0061] FIG. 35 is a diagram for explaining a functional model of a pinion gear in the gear unit shown in FIG. 33;

[0062] FIG. 36 is a diagram for explaining a functional model of the overall gear unit shown in FIG. 33;

[0063] FIG. 37 is a diagram showing a functional model of an engine accessory alone in FIGS. 37 to 42 that show the functional models of the entire power train system shown in FIG. 6;

[0064] FIG. 38 is a diagram showing a functional model of a torque converter alone in FIGS. 37 to 42 that show the functional models of the entire power train system shown in FIG. 6;

[0065] FIG. 39 is a diagram showing a functional model of a gear and differential alone in FIGS. 37 to 42 that show the functional models of the entire power train system shown in FIG. 6;

[0066] FIG. 40 is a diagram showing a functional model of an axle alone in FIGS. 37 to 42 that show the functional models of the entire power train system shown in FIG. 6;

[0067] FIG. 41 is a diagram showing a functional model of the travel load alone in FIGS. 37 to 42 that show the functional models of the entire power train system shown in FIG. 6;

[0068] FIG. 42 is a diagram showing a functional model of a planetary gear unit alone in FIGS. 37 to 42 that show the functional models of the entire power train system shown in FIG. 6;

[0069] FIG. 43 is a graph showing the simulation results of the power train shown in FIG. 6;

[0070] FIG. 44 is a graph showing the simulation results of the power train shown in FIG. 6;

[0071] FIG. 45 is a block diagram for explaining the hardware arrangement of a simulation apparatus according to the embodiment of the present invention;

[0072] FIG. 46 is a flow chart for explaining the processes in the simulation apparatus according to the embodiment of the present invention in turn; and

[0073] FIG. 47 is a chart for explaining the results obtained by the individual processes of the simulation apparatus according to the embodiment of the present invention.

[0074] A simulation system according to a preferred embodiment of the present invention will be described hereinafter with reference to the accompanying drawings.

[0075] <Potential Variable and Flow Variable>

[0076] The basic of unit expression as a characteristic feature of a simulation apparatus of this embodiment will be explained below.

[0077] Partial functions that built a system of a given entity are modeled as system elements which exchange energy. If a certain partial function is instantiated as a black box, as shown in FIG. 1, the energy model is defined by the internal loss of the black box, input energy to that block box, and output energy from that black box, and is given by:

[0078] Input energy=output energy+internal loss

[0079] The input energy and output energy are respectively expressed by state variables consisting of potential and flow variables. The potential variable is a state variable that represents the potential component of input/output energy with respect to the entity, and the flow variable is a state variable that represents the flow component of input/output energy with respect to the entity.

[0080] The state variable is expressed by a pair of input and output state variables. The input state variable is expressed by an input potential or flow variable, and the output state variable is expressed by an output potential or flow variable. That is, input/output energy is expressed by a pair of an input potential variable and an output flow variable or a pair of an input flow variable and an output potential variable.

[0081] Table 1 below summarizes an example of the potential and flow variables.

[0082] The energy model will be explained in more detail below. If a function of a motor is instantiated as a black box, and a current (value) is supplied to the flow variable of input energy, the potential variable of the input energy side is described as a voltage (value) corresponding to the internal resistance and output of the motor. In this case, the current is the input flow variable, the voltage is the output potential variable, and (current×voltage) is the input energy. More specifically, if the rotational torque remains the same, the voltage rises and the rotational speed of the motor increases as the current flows more. On the other hand, a torque is output as an output flow variable, and the angular velocity of the motor is returned as an input potential variable. In this case, the output energy is given by (torque ×angular velocity).

[0083] Another way of describing this model example is that a current is returned (output) as a flow variable by inputting a voltage as a potential variable. In this case, the arrows in FIG. 1 are reversed, i.e., the voltage is the input potential variable and the current is the output flow variable.

[0084] The interior of the black box that expresses the entity is constructed by a closed loop system (eigenvalue system) that defines the characteristics of that entity, and an open loop system that defines the characteristics of an environment which supplies input energy to the closed loop system, as shown in FIG. 2. In the closed loop system, as shown in FIG. 3, a permanent loop of the state variable is formed by feeding back the output state variable to the input state variable, and internal elements included in this loop settle an eigenvalue. In the open loop system, as shown in FIG. 4, the input state variable is applied to the output state variable to form a state variable flow, which settles the flow between an external state variable serving as input energy applied to the entity, and an internal state variable serving as input energy of the eigenvalue. Hence, an energy group applied to the entity applies potential and flow input energy groups to the environment system, and the potential and flow energy groups are applied to the eigenvalue system via the environment system.

[0085] In this way, all the partial functions that construct a product are expressed by energy models as system elements, which are combined to have a given connection relationship, thus modeling the product.

[0086] A virtual prototype to be described below is achieved on the basis of the aforementioned findings.

[0087] <Concept of Virtual Prototype>

[0088] FIG. 5 shows the conceptual arrangement of a virtual prototype.

[0089] In a model based on the virtual prototype, its contents are hierarchically classified into a “functional model” and “mechanism model”.

[0090] The functional model is described by the two generalized state variables, i.e., the “potential variable” and “flow variable” mentioned above. The “functional model” expresses the dynamic characteristics of an entity to be modeled, and plays a role of forming an integrated system by integrating subsystems. In the example of FIG. 5, the integrated system is built by subsystems (a) and (b).

[0091] The “mechanism model” gives physical characteristics to be substituted in parameters of the functional model, and depends on the mechanism of an entity.

[0092] The hierarchical model arrangement can model the overall function by integrating models of subsystems by a common method without troubling ourselves about handling of mechanism models which are normally complicated.

[0093] The function model is expressed by a system equation, and updates and discretizes parameters given by the mechanism model in response to each sampling clock. Hence, any nonlinearity of the characteristics resulting from a mechanism is linearized by the functional model at each sampling time.

[0094] <Basic Model of Power Train>

[0095] A power train includes elements such as a clutch, brake, and the like, which discontinuously switch model mechanisms by connecting/disconnecting models. Such structural change will be referred to as “structural nonlinearity” hereinafter.

[0096] Assuming a model that does not contain any structural nonlinearity, and an example of an arrangement constituted by functional and mechanism models, integration of subsystems into an integrated system, and solutions will be explained below.

[0097] The integrated model of the power train is constructed by four functional models, i.e., “engine & accessory”, “torque converter”, “gear box”, and “driving force”, as shown in FIG. 6. These functional models are coupled to mechanism models of the engine, torque converter, and the like via parameters. Note that the gear box is simplified to a reduction gear ratio alone.

[0098] Adjacent functional models are coupled via a torque and force as flow variables, and a rotation and velocity as potential variables.

[0099] <Driving Resistance Model of Engine & Accessory>

[0100] An engine functional model is expressed as a first-order delay system consisting of a moment J_engof inertia and viscous drag coefficient C_eng, as shown in FIG. 7.

[0101] An engine mechanism model gives the torque characteristics of the engine and the driving resistance of the accessory as the viscous drag coefficient and an offset value T_o.engas a frictional component of the drag to the functional model. The engine functional model contains an electronic control model that considers transient responses to a control input and driver's manipulation input, but does not consider any unstable phenomenon such as accidental fire.

[0102] A system equation of the engine functional model is given by: $\begin{array}{cc}{\left[\begin{array}{cc}{X}_{\mathrm{eng}}^{\prime }& O_{\mathrm{eng}}\end{array}\right]}^T=\left[\begin{array}{cc}{A}_{\mathrm{eng}}& B_{\mathrm{eng}}\\ C_{\mathrm{eng}}& D_{\mathrm{eng}}\end{array}\right]\times {\left[\begin{array}{cc}{X}_{\mathrm{eng}}& I_{\mathrm{eng}}\end{array}\right]}^T& \left(\mathrm{EQ}1\right)\end{array}$

[0103] Note that “′” represents a time differentiation in the above and subsequent equations.

[0104] Since X_engis a state vector and there is only one state variable X_eng, we have:

[0105] X_eng=[X_eng]  (EQ2)

[0106] When the functional model of the engine accessory is expressed by the block diagram shown in FIG. 8, the observation vector O^engand external input vector I_engfor extracting state variables are respectively given by:

[0107] O_eng=[ω_engT_accy]^TI_eng=[1T_eng]^T  (EQ3)

[0108] where T_engis the engine torque, ω_engis the rotational speed of the engine, and T_accyis the output torque to the accessory. Hence, parameters of the model that form a partial matrix are respectively given by: $\begin{array}{cc}{A}_{\mathrm{eng}}=\left[\frac{1}{J_{\mathrm{eng}}}\left(C_{\mathrm{eng}}-C_{\mathrm{accy}}\right)\right]& \left(\mathrm{EQ}4\right)\\ B_{\mathrm{eng}}=\left[\frac{1}{J_{\mathrm{eng}}}\left(T_{O.\mathrm{eng}}-T_{O.\mathrm{accy}}\right)-\frac{1}{J_{\mathrm{eng}}}\right]C_{\mathrm{eng}}=\left[\begin{array}{c}1\\ C_{\mathrm{accy}}\end{array}\right]& \left(\mathrm{EQ}5\right)\\ D_{\mathrm{eng}}=\left[\begin{array}{cc}0& 0\\ T_{O.\mathrm{accy}}& 0\end{array}\right]& \left(\mathrm{EQ}6\right)\end{array}$

[0109] Note that FIG. 9 shows symbols used in the system block diagrams of this embodiment as well as FIG. 8.

[0110] <Model of Torque Converter>

[0111] Upon modeling the torque converter, the torque converter is divided into a pump and turbine, and the former is rigidly coupled to the engine and the latter to the axle. Transient behavior of fluid in the torque converter is ignored.

[0112] The driving torque of the torque converter is given, using a capacitive coefficient C_cphaving a velocity ratio e between the pump and turbine as a variable, by:

[0113] T_eng=C_cp(e)·ω_eng²  (EQ7)

[0114] The mechanism model calculates a tangent to a driving torque curve shown in the graph in FIG. 10, and gives a slope C_tcand intercept T_O.tcto the functional model. At the same time, the mechanism model reads the torque ratio from characteristic data, and gives it to the functional model.

[0115] FIG. 10 is a block diagram of the functional model of the torque converter.

[0116] <Vehicle Driving Force Model>

[0117] The vehicle driving force model is constituted by a vehicle mass, travel resistance, tire radius, final reduction ratio, and moment of inertia of an axle shaft. The tire rolling resistance, air resistance, and hill-climbing resistance as factors of the travel resistance are extracted from the mechanism model, and are converted into a viscous drag coefficient and offset load to be given to the functional model.

[0118] Referring to FIG. 11, R_fis the final reduction gear ratio of the gear box, J_axis the inertial mass of the shaft, R_tyis the radius of the tire, M_vis the mass of the vehicle, C_vis the viscous drag coefficient that depends on velocity, and F_O.vis the offset force acting on the vehicle body.

[0119] <Integration of Functional Models>

[0120] The power train as an integrated unit of the aforementioned engine, torque converter, and driving force units (including the gear box) can be expressed by integrating their system equations.

[0121] FIGS. 12 to 14 are views for explaining integration of system equations. The integration is described in detail in Japanese Patent Application No. 7-250974 filed by the present inventors, and its contents will be quoted in the present specification.

[0122] More specifically, assume that there are three unit models to be integrated, as shown in FIG. 12. Then, parameter matrices of the individual units are placed at diagonal positions on a system equation, and input and output vectors are factored into state variables and input and observed state variables by manipulating the rows and columns of the matrix, as shown in FIG. 13.

[0123] As shown in FIG. 15, connected state variables (observed state variables) between two units are substituted in input state variables to delete the input state variables as connected state variables from input vectors. With such manipulations, the integrated system is expressed by an equation in the form of FIG. 14. As a result, the input vector I in this equation consists of only a constant term and, hence, is independent from time. In this way, the integrated model is completed.

[0124] The input vector I consists of only a constant term since offset amounts and the like are inserted in partial matrices A, B, and C of the individual unit models.

[0125] <Solution>

[0126] In order to simulate the integrated model, the integrated system equation must be solved.

[0127] In this embodiment, the integrated system equation is calculated in a discrete time system using zero-th-order hold. In a continuous-time system, this state equation is nonlinear, but in a discrete time system with a short unit time, the input I is expressed as a step input, i.e., the input I has a constant value within a short time Δt and, hence, it has zero-th order in terms of time. If conversion from the continuous-time system into the discrete-time system is done for a partial matrix [A B] to obtain a converted matrix [P Q], P and Q are, as shown in FIGS. 17 and 18, given by: $\begin{array}{cc}P=\exp \left(A·t_s\right)=E+A·t_s+\frac{A^2·t_s^2}{2!}+\cdots +\frac{A^n·t_s^n}{n!}& \left(\mathrm{EQ}8\right)\\ \begin{array}{c}Q={\int }_0^{t_s}B·\exp \left(A\eta \right)d\eta \\ =\left\{E+\frac{A·t_s}{2!}+L+\frac{A^n·t_s^n}{\left(n+1\right)!}+L\right\}·{\mathrm{Bt}}_s\end{array}& \left(\mathrm{EQ}9\right)\end{array}$

[0128] where E is the unit matrix, and t_Sis the sampling time. The system equation of the discrete system is given by: $\begin{array}{cc}{\left[\begin{array}{cc}{X}_{k+1}& O_k\end{array}\right]}^T=\left[\begin{array}{cc}P& Q\\ C& D\end{array}\right]\times {\left[\begin{array}{cc}{X}_k& I_K\end{array}\right]}^t& \left(\mathrm{EQ}10\right)\end{array}$

[0129] <Structurally nonlinear Power Train Functional Model>

[0130] A clutch is a typical mechanism element that controls structural nonlinearity for disconnecting or connecting systems. The power train to be simulated by this embodiment is an automatic gear box model having a ravigneaux planetary gear train as an example containing the nonlinear element.

[0131] Modeling of a clutch element upon handling such mechanical nonlinear clutch element will be explained below.

[0132] <Modeling of Clutch Element>

[0133] A clutch model is obtained on the basis of a mechanical impedance that receives a rotation difference between two systems connected thereto, and outputs a torque as a response.

[0134] As a mechanical impedance model, two models, i.e., a model that connects systems by rigidity and a model consisting of only a viscous coefficient, are available. Since the latter model is compared to a viscous coupling and is easy to handle although slippage slightly occurs upon engagement, the system of this embodiment as an automatic gear box model adopts the latter model consisting of only a viscous coefficient.

[0135] When the torque to be transmitted has exceeded a torque value that can be transmitted (to be referred to as a torque capacity T_Cehereinafter), the clutch begins to slip, and outputs a torque equal to the torque capacity T_Ce.

[0136] FIG. 19 shows the functional model of the clutch. The torque capacity T_Ceis given to the functional module by a clutch mechanism model. As shown in FIG. 19, the functional model calculates the torque value in an engaged state. More specifically, the rotational speed difference (=ω_C.in−ω_C.out) between the input and output is converted into a torque value T_inby a coefficient C_Ce, and the torque value is compared with the capacity T_Ce. If T_in>T_Ce, T_Ceis output; if T_in<T_Ce, T is output. More specifically, one with a smaller absolute value of T_inand T_Ceis selected.

[0137] A logic operation for selecting an output with a smaller absolute value is expressed by “switch”, as shown in FIG. 19. In order to reflect the logic operation result, in a system equation (to be described later) of the clutch, a switch coefficient S_Ewhich is set at “1” upon engagement or “0” upon slippage, and a switch coefficient S_Swhich assumes a value opposite to the coefficient S_Eare set. A parameter of the mechanical impedance portion is multiplied by the former (S_E), the torque capacity T_Ceis multiplied by the latter (S_S), and the sum of these products is output. The respective elements of the system equation of the clutch are: $\begin{array}{cc}{X}_C=0O_C={\left[\begin{array}{cc}{T}_{C.\mathrm{in}}& T_{C.\mathrm{out}}\end{array}\right]}^T{I}_C={\left[\begin{array}{ccc}{\omega }_{C.\mathrm{in}}& {\omega }_{C..\mathrm{out}}& 1\end{array}\right]}^T& \left(\mathrm{EQ}11\right)\\ A_C=B_C=C_C=0D_C=\left[\begin{array}{ccc}{S}_E{C}_{\mathrm{Ce}}& -S_E{C}_{\mathrm{Ce}}& S_S{T}_{\mathrm{Ce}}\\ S_E{C}_{\mathrm{Ce}}& -S_E{C}_{\mathrm{Ce}}& S_S{T}_{\mathrm{Ce}}\end{array}\right]& \left(\mathrm{EQ}12\right)\end{array}$

[0138] For, upon engagement,

[0139] S_E=1

[0140] S_S=0

[0141] upon slippage,

[0142] S_E=0

[0143] S_S=1

[0144] Note that a brake is handled as a clutch, one system of which is at rest.

[0145] More specifically, FIG. 20 is a block diagram of the functional model of the clutch (brake) as the viscous coupling. Referring to FIG. 20, C_-clutchis equivalent to C_Cein FIG. 19. Two switches S_-engageand S_-slipare respectively given by:

[0146] |T_in|≦|T_C|→S_-engage=1, S_-slip=0

[0147] |T_in|>|T_C|→S_-engage=0, S_-slip=1

[0148] <Functional Model of Automatic Gear Box>

[0149] The automatic gear box includes a plurality of clutches and brakes. The virtual prototype can model by placing a plurality of clutch and brake models in the same layout as the flow of power of the entity.

[0150] FIG. 21 shows the arrangement of the automatic gear box to be modeled by this embodiment. Referring to FIG. 21, reference numeral 1 denotes a torque converter; 2, a turbine shaft; 3, a forward (FWD) clutch; 4, a coaster clutch; 5, a 3-4 clutch; 6, a reverse clutch; 7, an L & R brake; 8, a 2-4 brake; and 9, a planetary gear train. FIG. 23 illustrates the power transmission in the automatic gear box shown in FIG. 21.

[0151] Referring to FIG. 23, I indicates the flow of power in the order of turbine shaft→FWD clutch→front sun gear, II indicates the flow in the order of turbine shaft→3-4 clutch→rear carrier (together with L & R clutches), III indicates the flow in the order of turbine shaft→reverse clutch→rear sun gear (together with 2/4 band brake), and IV indicates the flow in the order of front carrier→final gear.

[0152] FIG. 24 is a diagram obtained by directly modeling the connections among the building elements shown in FIG. 23. Referring to FIG. 24, the rectangular symbol indicates a transfer function (mechanical impedance or mobility), the broken arrow indicates the flow of the potential variable (angular velocity), and the solid allow indicates the flow of the flow variable (torque). Note that this diagram is premised on power=torque×angular velocity.

[0153] The turbine shaft is modeled according to FIG. 25. More specifically, the turbine shaft is handled as a particle system model of a first-order delay system consisting of an inertia (J-trb) and viscous drag (D-trb). FIG. 26 is a block diagram of this model. In FIG. 26, a torque T-tc is a torque input from the torque converter. A state equation of the turbine shaft according to this block diagram is: $\begin{array}{cc}{x}^{\prime }=-\frac{\mathrm{Dtrb}}{\mathrm{Jtrb}}·x+\frac{\mathrm{Ttc}-\mathrm{Ttrk}}{\mathrm{Jtrb}}& \left(\mathrm{EQ}13\right)\end{array}$

[0154] FIG. 27 shows the response characteristics of this turbine shaft according to the aforementioned state equation. FIG. 27 shows that the model has first-order response delay characteristics.

[0155] Modeling of the FWD clutch, 3-4 clutch, reverse clutch, L & R clutches, and 2-4 band brake will be examined. A general clutch has been explained above with reference to FIG. 19. In the following description, various clutches and brakes unique to the automatic gear box will be examined.

[0156] Initially, the coaster clutch that is not directly associated with transmission can be excluded from objects to be modeled.

[0157] FIG. 28 shows the functional model of a clutch. Referring to FIG. 28, μ is a coefficient of friction, R_Eis the effective diameter of the clutch, T_Cis the torque capacity, and FC is the pressing force. Hence, the torque capacity T_Cis given by:

[0158] T_C=μ·R_E·F_C  (EQ14)

[0159] The block diagram in FIG. 20 is assigned to the model shown in FIG. 28.

[0160] In a one-way clutch, the engaged and non-engaged states are switched in correspondence with the rotational direction. This operation can be interpreted that the torque capacity T_Ceassures zero or a value large enough to ensure engagement in correspondence with the rotational direction. That is,

[0161] T_C=0:

[0162] when rotational direction=forward direction,

[0163] T_C=T_o(for T₀>>1):

[0164] when rotational direction=reverse direction. Hence, when the torque capacity that considers layout and operation of a frictional clutch and one-way clutch is given, and these clutches are expressed by a single model, the characteristics shown in FIG. 29 are obtained.

[0165] In this gear box, the FWD clutch and one-way clutch (OWC) are laid out in series with each other. In this series layout, if either the FWD clutch or one-way clutch is not engaged, slippage occurs. More specifically, a clutch having a torque capacity with a smaller absolute value is selected. In other words, the FWD clutch and one-way clutch (OWC) can be replaced by a single clutch model, which selects a clutch having a torque capacity with a smaller absolute value.

[0166] On the other hand, the L & R brakes select a brake having a torque capacity with a larger absolute value since only one brake can be engaged due to parallel layout.

[0167] The mechanism model obtains the torque capacity from the corresponding clutch/brake hydraulic pressure according to the operation of the control system of the automatic gear box, and gives it to the functional model.

[0168] This portion uses an existing simulation program. The viscous coefficient values such as C_Ce.Pand the like which functions in an engaged state are set in consideration of slippage upon engagement and spike-like torque generation at the instance of engagement.

[0169] <Functional Model of Planetary Gear Train>

[0170] The automatic gear box of this power train is designed to have a planetary gear box. In this embodiment, the functional model of a ravigneaux planetary gear train is proposed as the planetary gear train.

[0171] FIGS. 30 and 31 show the ravigneaux planetary gear train respectively in terms of rotation (ω) and torque transmission (T).

[0172] This planetary gear train is modeled paying attention to the motion of a long pinion. In this modeling, the rotations of the input and output shafts of the planetary gear train are expressed using:

[0173] I: rotation of a ring gear as the output shaft; and

[0174] II: rotation of a pinion gear.

[0175] With reference to FIG. 30, the rotations of the respective shafts, i.e., a rotational velocity ω_SSof a small sun gear, a rotational velocity ω^Cof a carrier, and a rotational velocity ω^LSLS of an E sun gear are respectively given by:

[0176] ω_SS=G_SSω_LP+ω_R

[0177] ω_C=G_Cω_LP+ω_R

[0178] ω_LS=G_LSω_LP+ω_R  (EQ15)

[0179] where G_SS, G_C, and G_LSare the gear ratios between the pinion and the respective shafts on the condition that the ring gear is fixed in position, which include revolution upon rotation of the pinion, and are respectively defined by: $\begin{array}{cc}{G}_C=-\frac{z_{\mathrm{LP}}}{z_R},G_{\mathrm{SS}}=\frac{z_{\mathrm{LP}}}{z_R}\left(\frac{z_R}{z_{\mathrm{SS}}}-1\right),G_{\mathrm{LS}}=-\frac{z_{\mathrm{LP}}}{z_R}\left(\frac{z_R}{z_{\mathrm{LS}}}+1\right)& \left(\mathrm{EQ}16\right)\end{array}$

[0180] Torque balance will be examined below with reference to FIG. 31. Let T_SS, T_C, and T_LSbe the torque components from the respective input shafts to the planetary gear system, and T_Rbe the output torque. For the sake of simplicity, the inertia and resistance of each shaft are ignored. To balance the entire system, since we have:

[0181] T_R=T_SS+T_C+T_LS  (EQ17)

[0182] the equation of motion around the pinion shaft is:

[0183] 1/J_LPω′_LP+D_LPω_LP=T_LP.R+T_LP.SS+T_LP.LS=G_SST_SS+G_CT_C+G_LST_LS  (EQ18)

[0184] where “′” indicates a time differentiation.

[0185] FIG. 32 is a block diagram of this ravigneaux planetary gear train modeled in this way.

[0186] The characteristic feature of the block diagram of FIG. 32 is that a circuit portion that receives/outputs T_LPand ω_LPexpresses rotation of the long pinion, and other circuit portions express power transmission balance of revolution components. In other words, rotation “interferes with” revolution via the gear ratios. That is, rotation and revolution are expressed in parallel with each other. If rotation and revolution are expressed in series with each other, modeling must be redone every time the conditions and states change. However, since parallel expression is used, the planetary gear train can be expressed by changing only parameters even when the conditions or states change.

[0187] The advantage of the parallel expression of the planetary gear train is evident when such expression is applied to gear mechanisms other than the ravigneaux gear train.

[0188] A function model when the planetary gear train is a CR—CR planetary gear train will be examined below.

[0189] FIG. 33 shows the structure of the CR—CR planetary gear train. This gear train has a pinion gear which revolves around a sun gear between the sun gear and a ring gear. In this CR—CR planetary gear train, a rear ring gear and front pinion gear are coupled via a planetary carrier, and the front pinion gear and a rear pinion gear are coupled via a planetary carrier. FIG. 34 shows the state wherein a pinion gear revolves around a sun gear, and a carrier rotates around a sun gear shaft together with revolution of a pinion gear. The ring gear meshes with the pinion gear.

[0190] This planetary gear train is modeled in regard to the motion of the long pinion. This modeling expresses the rotations of the input and output shafts of the planetary gear train by rotations I and II in FIG. 34, i.e.:

[0191] I: rotation of the ring gear as the output shaft (i.e., the rotation of the overall system); and

[0192] II: rotation of the pinion gear.

[0193] In this case, assume that the output shaft (front carrier) is fixed in position. Referring to FIG. 34, let J-fp be the inertia of the front pinion, D-fp be the viscous coefficient of the front pinion, and ω-fp be the angular velocity of the front pinion. Then, the gear train shown in FIG. 34 is modeled, as shown in the block diagram in FIG. 35. Hence, the entire planetary gear train is modeled, as shown in the block diagram of FIG. 36.

[0194] As can be seen from the block diagram in FIG. 36, even in the CR—CR gear mechanism, contribution components of rotation of the front pinion gear, and those of revolution of the entire system are parallelly separated.

[0195] <Integration to Integrated Model>

[0196] The gear box model simply expressed by coupling of only reduction ratios is replaced by an integrated functional model of the gear box mechanism (FIG. 24) and the planetary gear train (FIGS. 30 and 31) to built an integrated model. This integration uses the scheme that has been explained earlier in relation to FIGS. 12 to 15.

[0197] FIGS. 37 to 42 are block diagrams of the entire power train to be simulated, which is applied to the simulation apparatus of this embodiment. Of these drawings, FIGS. 37 to 41 are joined in series with each other, and the block diagram of the planetary gear mechanism shown in FIG. 42 is built in the gear block shown in FIG. 39, thus completing the overall block diagram.

[0198] <Simulation Result>

[0199] The matrix expression of the integrated model obtained in this way is simulated along the time axis of a discrete system. A simulation sample including nonlinear parameters will be explained.

[0200] More specifically, FIG. 43 shows the simulation result of variations of engine rotation in response to the load on an engine accessory input to an idling vehicle at a given timing. Upon comparing the simulation result (solid curves) and test results (broken curves) using an actual vehicle in FIG. 43, modeling by the simulation apparatus of this embodiment can provide sufficiently high simulation precision levels when engine rotation and behavior of the engine control system are to be expressed. In the simulation result, rotation recovery is early, and there is no vibration after the rotational speed has risen, unlike the actual vehicle tests. This is because the simulation considers neither instability of combustion nor any vibration system such as an engine mount or the like.

[0201] FIG. 44 shows a simulation sample containing structural nonlinearity.

[0202] FIG. 44 shows the simulation result of shift changes of the automatic gear box obtained in response to throttle manipulation during travel, using the simulation model shown in FIG. 3. Actually measured values are measurement data obtained by reproducing an identical situation on a test bench. The simulation result well matches the test result of the actual vehicle.

[0203] In this way, the engine, automatic gear box, vehicle driving load, and control system, which are modeled individually on the basis of the concept of the virtual prototype are integrated to develop a model of the overall power train. An examination of the model revealed that sufficiently high precision can be obtained, thus proving effectiveness of modeling based on the virtual prototype.

[0204] <System Arrangement>

[0205] FIG. 45 shows the hardware arrangement of the simulation apparatus according to this embodiment. This hardware comprises a CPU 1, display 2, keyboard 3, mouse 4, main memory 5, file disk 6, and file disk 7 for processing software.

[0206] The flows of data and files in the simulation apparatus of FIG. 45 will be explained with reference to FIG. 46.

[0207] Referring to FIG. 46, a functional model file stores a functional model of each unit generated in advance as data. Generation of the functional model file is described in detail in Japanese Patent Application No. 7-250974 filed by the present applicant. Also, in FIG. 46, a mechanism model file 1100 stores data corresponding to the aforementioned “mechanism model” which may be implemented as software to run in a system according to FIG. 45, and a data model file 1300 stores, static characteristic data, e.g., physical characteristic data as functions or tables, as shown in FIG. 10, respectively for each unit (step S1).

[0208] The system of this embodiment registers as a unit model 1200 (step S2) each functional model (each functional model represented/defined as a system equation) and the mechanism model corresponding to each functional model. Then the system equations of each unit model 1200 can be integrated by the aforementioned method (step S3). This integration uses deletion of connected state variables, and generates an integrated system equation (step S3). The generated integrated system equation representing/defining the target apparatus can be stored in an integrated system equation file 1400 (step S4).

[0209] Simulation is done according to the integrated system equation (step S5). In the simulation, static characteristics are read out from the data model and provided to the “mechanism model”, and are converted by the “mechanism model” into parameters to be supplied to the integrated system equation in each sampling clock. The integrated system equation is calculated/solved along a discrete time axis according to the aforementioned method, which may be referred to as discretization. The simulation progresses while repeating supply of parameters and discretization.

[0210] The aforementioned simulation outputs, for example, on display 4, a simulation result as shown in FIGS. 43 and 44 (step S6).

[0211] FIG. 47 shows the overall simulation system completed by the method of the present invention, and illustrates that an integrated apparatus can be simulated efficiently based on the individual units of the vehicle.

[0212] As described above, according to the present invention, a plurality of kinds of units can be easily evaluated without changing models.