Method of and System for Avoiding Collisions Between Aircraft

My present invention relates to a method of and a system for preventing in-flight aircraft collisions by the evaluation of information received aboard a given aircraft from nearby aircraft periodically transmitting navigational data relating to their course, speed, altitude and so on.

In another copending application, Ser. No. 686,657 filed May 2, 1976, now abandoned, I have disclosed an integrated digital inter-aircraft communication and location system. Insofar as relevant to the present invention, that system individually allots to each aircraft in a given area a recurrent time slot for the emission of their navigational messages, with different carrier frequencies assigned to aircraft transmitting in consecutive time slots. The several carrier frequencies define a recurrent frame composed of a predetermined number of time slots which need not coincide with the number of intercommunicating aircraft; these time slots are established by pulses periodically emitted from an associated ground control station.

Aboard a given aircraft, the apparent velocity of another aircraft--i.e. the speed with which the two craft approach or recede from each other--can be determined on the basis of the Doppler shift, i.e. the deviation of the received carrier frequency from the nominal value of that frequency. If that Doppler shift is positive, the approach speed V.sub.d can be readily calculated. If the two craft are found to approach each other and are on the same altitude level, a possible collision course may be presumed. From the available speed and heading information, the true relative velocity can be vectorially ascertained. If the two craft are actually on a collision course, this relative velocity V.sub.r corresponds to the approach speed V.sub.d ; otherwise, V.sub.d =V.sub.r cos.alpha. where .alpha. is the angle included between the two vectors as further discussed hereinafter (at large inter-aircraft distances, however, this angle approaches zero). From the delay .DELTA.t between the arrival of a timing pulse from the ground control station and the reception of a message from the communicating aircraft to which the time slot initiated by this pulse is assigned, the first aircraft can obtain information on the minimum distance separating the two craft, that minimum distance being equal to c.DELTA.t/2 where c is the propagation velocity of the emitted radio waves. The actual distance may be considerably greater, depending upon the relative positions of the two aircraft and the ground station.

With these data, conventional arithmetic equipment aboard each aircraft can calculate a minimum closing time remaining up to the instant of closest approach; the mathematical basis for this calculation will be presented below. Such calculation, however, is valid only in first approximation, not only because of the limited reliability of the Doppler-shift evaluation (due to possible fluctuations in carrier frequency) but also in view of the aforementioned uncertainty in the determination of distance. Thus, an undue number of false alarms would result if a dropping of this minimum closing time below a certain safety threshold were taken as a signal to the crew of an aircraft to take collision-avoidance measures.

The general object of my present invention, therefore, is to provide a method of avoiding such false alarms in an aerial-navigation system of the type referred to.

A related object is to provide means in such a system for carrying out this method.

In accordance with my present invention, the detection aboard a given aircraft of a presumed flight path of another aircraft crossing its own flight path with a minimum closing time less than a predetermined safety threshold does not immediately trigger an alarm but, instead, initiates a direct exchange of messages at an accelerated rate between the two aircraft. From this direct exchange, the initiating aircraft determines the distance D separating the two aircraft and the variation .DELTA.D of that distance from one exchange cycle to the next. This distance variation .DELTA.D is proportional to the apparent speed V of the other aircraft as seen from the initiating aircraft and is used, together with distance D, for calculating a passing distance s. Collision-avoiding measures are taken only when this passing distance s is found to be less than a predetermined magnitude.

The switchover from the evaluation of the Doppler shift to a more accurate speed determination by a repeated calculation of the actual inter-aircraft distance D, upon the initiation of the high-speed exchange, provides a considerably more reliable measure of the closing time .tau. than any conventional procedure, including the one outlined above. Moreover, the passing distance s can be computed almost immediately upon that switchover, thus when the closing time is still relatively long, to apprise the pilot of the possible need for changing course or velocity in order to avoid a collision. These calculations do not involve the ground control station and are also independent of possible fluctuations in the carrier frequency transmitted by the responding aircraft.

The above and other features of my invention will now be described in detail with reference to the accompanying drawing in which:

FIG. 1A is a graph showing the periodic emission of messages from several aircraft cruising in a certain area within reach of a ground control station;

FIG. 1B is a diagrammatic representation of a message sent out from such aircraft;

FIG. 2 is a graph showing an accelerated exchange of messages between two aircraft on a potential collision course;

FIG. 3A is a vector diagram showing the flight paths and absolute velocities of the two aircraft;

FIG. 3B is a vector diagram showing the relative position of the two aircraft at a certain point of their approach;

FIG. 3C is a graph showing the relative velocity of the two aircraft over a period including the instant of closest approach;

FIG. 3D is a graph showing the relative acceleration of the two aircraft over the same period;

FIG. 4 is a block diagram of radiocommunication equipment on board an aircraft for emitting and receiving the messages of FIGS. 1A, 1B and 2; and

FIG. 5 is a partial block diagram of a processor forming part of the equipment of FIG. 4.

As shown in FIG. 1A, a number of aircraft A.sub.1, A.sub.2, A.sub.3, A.sub.4, A.sub.5. . . cruising within reach of a common ground control station periodically emit respective messages on frequencies f.sub.1, f.sub.2, f.sub.3 and f.sub.4 individually assigned to them. The ground station broadcasts timing pulses or spikes at equispaced instants t.sub.1, t.sub.2, t.sub.3, t.sub.4, defining respective time slots in a recurrent frame of four such time slots; these spikes, of course, are received by the several aircraft at different times depending upon their distance from the transmitter of the control station. The duration T of each time slot, however, is sufficient to prevent any overlap between the recurrent messages emitted by different aircraft. Frequencies f.sub.1 -f.sub.4 may each be used by more than one aircraft, as specifically illustrated for craft A.sub.1 and A.sub.5.

In FIG. 1B I have shown a possible structure of such a recurrent message. It comprises an acquisition section A-B, a carrier-emission section B-C, a section C-D for the broadcasting of an identification code, a section D-E for sending out a reception code, a section E-F giving altitude information, a section F-G containing heading information, a section G-H reporting changes in altitude, and a section H-I for transmitting a synchronizing pulse. The balance I-J of the transmitting interval is reserved for propagation time.

From the altitude and heading information, as well as from the Doppler deviation of the carrier frequency transmitted in section B-C, each aircraft within range of the ground station can readily ascertain--by manual plotting or with the aid of a computer--whether any other aircraft is on a possible collision course with it and, if so, whether the presumed minimum closing time is less than a predetermined threshold. In the latter case it switches to a rapid exchange of information with that other aircraft as illustrated in FIG. 2; the responding aircraft then sends out its own message immediately upon receiving the synchronizing pulse H-I from the inquiring aircraft so that the delay between incoming and outgoing sync pulses is a measure of the distance between the two aircraft, taking into account the length of the message. This exchange of information may be carried out on their respective frequencies (f.sub.1 and f.sub.2 in the case of aircraft A.sub.1 and A.sub.2, as represented in FIG. 2) with a modified reception code, for example, to indicate this stage of communication; alternatively, a separate frequency channel may be utilized for this purpose. Naturally, the messages thus transmitted must also carry a code identifying the other aircraft involved in the exchange; reference may be made in this connection to U.S. Pat. No. 3,341,845, for example.

In FIG. 3A I have shown the two aircraft A.sub.1, A.sub.2 headed toward an intersection X of their respective flight paths FP.sub.1 and FP.sub.2 at speeds V.sub.1 and V.sub.2, it being assumed that each craft travels at a substantially constant velocity and on a straight line. The relative velocity of these aircraft is indicated at V.sub.r and is considered invariant over the period under discussion. Successive positions of aircraft A.sub.1, reached after a series of equal time intervals, are shown at A.sub.1.sup.vi, A.sub.1 '", A.sub.1 ", A.sub.1 ' and A.sub.1 .degree.; corresponding positions of aircraft A.sub.2 have been designated A.sub.2.sup.iv, A.sub.2 '", A.sub.2 ", A.sub.2 ' and A.sub.2 .degree.. In these positions the two aircraft are separated by progressively decreasing distances D.sup.iv, D'", D", D' and D.degree.=s, the latter being known as the passing distance and marking the closest approach of the craft to each other. The angle included between these distances and the vector V.sub.r, designated .alpha..sup.iv, .alpha."', .alpha.", .alpha.' and .alpha..degree. for the several positions referred to, increases progressively from near-zero (with the two craft widely separated) to 90.degree. at the instant of closest approach. The passing distance s is seen to be perpendicular to the relative velocity V.sub.r.

FIG. 3B represents one of the triangles of FIG. 3A defined by the inter-aircraft distance D, the passing distance s and the speed vector V.sub.r including an angle .alpha. with line D. This distance D changes at a rate dD/dt=V, with the distance variation or apparent speed V given by

The point of closest approach (or of collision with s=0), as seen from aircraft A.sub.1, has been indicated at P.

Passing distance s satisfies the relationship

The apparent speed V has been plotted against time t in FIG. 3C. As can be seen from that graph, V goes through zero for D=s.

The apparent acceleration .gamma.=dV/dt, plotted against time t in FIG. 3D, has its maximum absolute value .gamma..sub.max at D-s. Its value is given by

With aircraft A.sub.2 moving at speed V.sub.r toward the point of closest approach P as seen from aircraft A.sub.2 (FIG. 3B), the closing time .tau. is given by ##EQU1## If the two craft were in fact on a collision course, that closing time would reduce (with s=0) to

(since V.sub.r =V under these circumstances).

With the minimum relative distance D and the approximate approach velocity V known aboard aircraft A.sub.1 from the normal radiocommunication described with reference to FIGS. 1A and 1B, a tentative closing time .tau. in conformity with equation (5) can be calculated from these data. In certain instances the closing time .tau. may be slightly shorter, as per equation (5), but only if s>0, i.e. if there is no real danger of collision.

When this tentative closing time drops below a predetermined safety threshold, a switchover to the rapid exchange according to FIG. 2 permits a more exact evaluation, without reliance on the Doppler shift, of distance D and, with it, of its derivatives V and .gamma.. An extrapolation of the acceleration function .gamma.(t) to determine its maximum yields a projected value for the passing distance s, with substitution of V for V.sub.r in equation (3). Since V<V.sub.r in a noncollision situation, the actual passing distance will be larger in any case in which .gamma..sub.max has a finite value so that the error lies on the side of safety.

If and when this calculated passing distance s is found to be less than a predetermined magnitude, the pilot of aircraft A.sub.1 will take collision-avoiding action such as a change in course and/or velocity. As long as these maneuvers do not involve a change in altitude, the ground-station controller need not intervene.

Obviously, the initial distance determination could also be based upon positional data if such information were included in the message transmitted by each aircraft.

In FIG. 4 I have illustrated, by way of example, certain on-board equipment that can be used to carry out the aforedescribed method of collision avoidance. The equipment comprises a coupler 10 with an array of antennas 11, 12 connected by it to a duplexer 13. The latter has an input alternately connectable by a switch 14 to a main transmitter 15 and an ancillary transmitter 16 which are energizable via a transformer 17 and pilot amplifiers 18, 19 as well as a switch 20 (ganged with switch 14) from an interface unit 21 including means for storing the data to be transmitted. Unit 21 co-operates with a processor 22 which obtains the information on altitude, heading etc. from nonillustrated measuring devices. Duplexer 13 also feeds incoming messages via a switch 23 and two alternately operable amplifiers 24, 25 to a set of receivers 26, 27, 28, 29 tuned to the several frequency channels f.sub.1 -f.sub.4 used in the system. Through a multiplexer 30, controlled by processor 22, these receivers work into the storage circuits of unit 21.

Incoming timing pulses from the ground control station are relayed by memory 21 to processor 22 for establishing the several time slots t.sub.1 -t.sub.4 described with reference to FIG. 1A. These timing pulses serve to synchronize, via a lead 31, a clock circuit 32 which delivers clock pulses via a lead 33 to the memory 21 and to the other circuit components.

To the extent relevant in conjunction with my present invention, processor 22 may comprise the units shown in FIG. 5, including a message analyzer 34 normally determining minimum distance D and Doppler speed V.sub.d, a calculator 35 operating on these parameters in accordance with equation (5) to emit an output signal whenever the minimum closing time tentatively determined during normal radiocommunication falls below its safety threshold, an electronic switch 36 closed by this output signal to feed the parameter D--now derived from an accelerated message exchange--to a first differentiator constituted by a comparator 37 with a subtractive input directly connected to switch 36 and an additive input connected thereto through a delay line 38, a second differentiator constituted by a comparator 39 with a subtractive input directly connected to comparator 37 and an additive input connected thereto through a delay line 40, and an extrapolator 41; all these units may operate digitally. The delay periods of lines 38 and 40 correspond to an exchange (interrogation/response) cycle as shown in FIG. 2.

Differentiator 37, 38 produces a signal representing the apparent velocity V derived from the distance decrement .DELTA.D; differentiator 39, 40 analogously generates a signal representing the apparent negative acceleration .gamma. derived from the velocity decrement .DELTA.V. Extrapolator 41 establishes the projected maximum deceleration .gamma..sub.max on the basis of successively obtained values of .gamma.. Another calculator 42 operates on the parameters V and .gamma..sub.max, pursuant to equation (3), to determine the passing distance s and to trigger an alarm device 43 if its value is less than the predetermined minimum. The output signal of calculator 35 also switches a message former 44 and a control circuit 45 for multiplexer 30 (FIG. 4) upon a changeover to the high-speed operation described in connection with FIG. 2; the operation of message analyzer 34 is concurrently modified, with a switchover from Doppler-shift evaluation to direct determination of distance D from the timing of the received response.

With a duration of 1.5 ms for each time slot t.sub.1 -t.sub.2 etc, and thus of 6 ms for a frame period R, the message transmission from one aircraft to the other as shown in FIG. 2 may last for about 2 ms. Within an interval of 10 frames, or 60 ms, 15 exchange cycles may therefore take place between the intercommunicating aircraft, sufficient to allow for a determination of passing distance s as described above.