Target prediction and collision warning system

• US 5,638,281 A
• Filed: 06/20/1995
• Issued: 06/10/1997
• Est. Priority Date: 01/31/1991
• Status: Expired due to Term

• Current Assignee: AIL Systems, Inc. (ITT, Inc.)
• Original Assignee: AIL Systems, Inc. (ITT, Inc.)
• Inventors: Wang, C. David
• Primary Examiner(s): ZANELLI, MICHAEL J
• Application Number:
  US08/492,895
• Time in Patent Office: 721 Days
• Field of Search: 364/461, 364/516, 342/70, 342/71, 342/72, 342/29, 342/41, 342/82, 342/83, 180/167, 180/169, 395/90, 395/902, 395/905, 340/961, 340/436
• US Class Current: 701/301
• CPC Class Codes: G01J 3/453 by correlation of the amplitudes
• CPC Class Codes: G01S 13/24 using frequency agility of carrier wave
• CPC Class Codes: G01S 13/348 using square or rectangular modulation, e.g. diplex radar for ranging over short distances
• CPC Class Codes: G01S 13/726 Multiple target tracking
• CPC Class Codes: G01S 13/931 of land vehicles
• CPC Class Codes: G01S 17/95 for meteorological use
• CPC Class Codes: G01S 2013/93271 in the front of the vehicles
• CPC Class Codes: G01S 2013/93272 in the back of the vehicles
• CPC Class Codes: G01S 2013/93274 on the side of the vehicles
• CPC Class Codes: G01S 7/021 Auxiliary means for detecting or identifying radar signals or the like, e.g. radar jamming signals
• CPC Class Codes: G01S 7/2883 using FFT processing
• CPC Class Codes: G01S 7/2921 based on data belonging to one radar period
• CPC Class Codes: G01S 7/417 involving the use of neural networks
• CPC Class Codes: G06N 3/047 Probabilistic or stochastic networks
• CPC Class Codes: G06N 3/049 Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
• CPC Class Codes: Y02A 90/10 Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
• CPC Class Codes: Y10S 706/905 Vehicle or aerospace

Abstract

A device for target prediction and collision warning for tracking objects in a region proximate to a vehicle includes a signal transmitter which provides first and second detection signals for at least partial reflection by an object located in a spatial region. The device further includes a signal receiver for receiving the deflected first and second detection signals corresponding to first and second parameter signals. A Fourier transform circuit is provided for receiving the first and second object parameter signals and generating first and second Fourier transform object parameter signals corresponding to relative range and velocity data of a target being tracked. The device includes a probabilistic neural network which preferably sorts the first and second Fourier transform object parameter signals corresponding to the relative range and velocity of a target being tracked. Operatively coupled to the probabilistic neural network is a target tracker circuit which receives the sorted first and second Fourier transform object parameter signals after at least three samples of relative range and velocity data have been measured. The target tracker generates an output signal indicative of a prediction of regression parameters of a second order or higher order equation that characterizes the change in relative range and velocity of the target being tracked.

Claims

1. A target prediction and collision warning system for detecting targets in a spatial region about a vehicle, the system comprising:

a) a signal transmitter providing at least first and second detection signals to the spatial region, the at least first and second detection signals at least partially reflecting off at least one object within the spatial region thereby generating at least partially reflected first and second detection signals, each of the at least partially reflected first and second detection signals being indicative of at least one characteristic of the at least one object within the spatial region;

b) a signal receiver operatively coupled to the signal transmitter, the signal receiver receiving each of the at least partially reflected first and second detection signals, each of the at least partially reflected first and second detection signals corresponding to first and second object parameter signals;

c) a target tracker circuit operatively coupled to the signal receiver, the target tracker circuit being responsive to at least one of the first and second object parameter signals and providing a regression parameter signal indicative of at least one of a past, current and future value of the at least one characteristic of the at least one object within the spatial region whereby the present and future position of a target within a spatial region may be predicted and imminent collision with the vehicle indicated, the regression parameter signal comprising regression coefficients that fit a second or higher order polynomial, whereby through data extrapolation the at least one of a past, current and future value of the at least one characteristic may be recreated; and

d) a data recorder operatively coupled to the target tracker circuit, the data recorder receiving and storing the coefficients for future evaluation whereby post-collision reconstruction of the targets'"'"' relative trajectory may be facilitated.

2. A target prediction and collision warning system as defined by claim 1, further comprising a probabilistic neural network operatively coupled between the signal receiver and the target tracker, the probabilistic neural network receiving the first and second object parameter signals and providing a probabilistic neural network output signal indicative of the at least one characteristic of the at least one object within the spatial region.

3. A target prediction and collision warning system as defined by claim 2, wherein the probabilistic neural network comprises:

a) a plurality of cluster processor circuits receiving the first and second object parameter signals, each cluster processor circuit generating an output signal representing a probability density function estimation value corresponding to the first and second object parameter signals, each cluster processor circuit including;

1) an input buffer memory circuit having a plurality of serially connected registers for storing the first and second object parameter signals assigned to a respective cluster processor circuit;

2) a plurality of processing elements, each of the plurality of processing elements being operatively coupled to a respective register of the input buffer memory circuit being responsive to assigned first and second object parameter signals stored in the input buffer memory circuit, each of the processing elements further being responsive to current unassigned first and second object parameter signals and providing a processing element output signal in response thereto;

3) a plurality of exponential function circuits, each of the plurality of exponential function circuits being coupled to a corresponding processing element, each of the plurality of exponential function circuits performing an exponential function on the processing element output signal and providing an exponential function circuit output signal in response thereto; and

4) a combiner circuit coupled to each of the plurality of exponential function circuits of the respective cluster processor circuit, the combiner circuit being responsive to the exponential function circuit output signals, performing an addition function thereon and providing an output signal representing a probability density function estimation value for each unassigned first and second object parameter signal;

b) a decision logic circuit operatively coupled to the combiner circuit of each cluster processor circuit, the decision logic circuit comparing the combiner circuit output signal of the corresponding cluster processor circuit with at least a first threshold value signal, and providing a decision address signal in response thereto; and

c) a switching circuit operatively coupled to the decision logic circuit and to each of the cluster processor circuits, the switching circuit further being responsive to current unassigned first and second object parameter signals and assigning the current unassigned first and second object parameter signals to the input buffer memory circuit of the respective cluster processor circuit in response to the decision address signal provided the decision logic circuit.

4. A target prediction and collision warning system for detecting targets in a spatial region about a vehicle, the system comprising:

a signal transmitter providing at least first and second detection signals to the spatial region, the at least first and second detection signals at least partially reflecting off at least one object within the spatial region thereby generating at least partially reflected first and second detection signals, each of the at least partially reflected first and second detection signals being indicative of at least one characteristic of the at least one object within the spatial region;

a signal receiver operatively coupled to the signal transmitter, the signal receiver receiving each of the at least partially reflected first and second detection signals, each of the at least partially reflected first and second detection signals corresponding to first and second object parameter signals; and

a target tracker circuit operatively coupled to the signal receiver, the target tracker circuit comprises a Kalman filter circuit, the Kalman filter circuit further comprising;

a recursive vector determination circuit operatively coupled to the signal receiver, the recursive vector determination circuit being responsive to the second object parameter signal and generating a recursive vector determination circuit output signal in response thereto;

a first recursive mixer circuit operatively coupled to the recursive vector determination circuit, the first recursive mixer circuit being responsive to the recursive vector determination circuit output signal and regression parameter signal, the first recursive mixer circuit effectively combining the recursive vector determination circuit output signal and an initial regression parameter signal and providing a first recursive mixer circuit output signal in response thereto;

a first recursive combiner circuit operatively coupled to the first recursive mixer circuit and being responsive to the first recursive mixer circuit output signal, the first recursive combiner circuit also being operatively coupled to the signal receiver and being responsive to the first object parameter signal, the first recursive combiner circuit effectively combining the first recursive mixer circuit output signal and the first object parameter signal and providing a first recursive combiner circuit output signal in response thereto;

a processing circuit operatively coupled to the recursive vector determination circuit and being responsive to the recursive vector determination circuit output signal and providing a processing circuit output signal in response thereto;

a second recursive mixer circuit operatively coupled to the first recursive combiner circuit and the processing circuit, the second recursive mixer circuit being responsive to and combining the first combiner circuit output signal and the processing circuit output signal, the second mixer circuit providing an output signal in response thereto; and

a second recursive combiner circuit operatively coupled to the second recursive mixer circuit, the second recursive combiner circuit being responsive to and combining the second recursive mixer circuit output signal and the regression parameter signal and providing an output signal indicative of an updated regression parameter signal corresponding to an output signal of the Kalman filter circuit, the Kalman filter circuit being responsive to at least one of the first and second object parameter signals and providing a regression parameter signal indicative of at least one of a past, current and future value of the at least one characteristic of the at least one object within the spatial region whereby the present and future position of a target within a spatial region may be predicted and imminent collision with the vehicle indicated.

5. A target prediction and collision warning system as defined by claim 4 which further comprises:

a delay circuit operatively coupled to the output port of the second recursive combiner circuit and receiving an updated regression parameter signal and generating the regression parameter signal, the delay circuit also being operatively coupled to the first recursive mixer circuit, the delay circuit providing the regression parameter signal to the first recursive mixer circuit and the second recursive combiner circuit.

6. A target prediction and collision warning system as defined by claim 4, wherein the processing circuit comprises:

a recursive scaler determination circuit operatively coupled to the recursive vector determination circuit, the recursive scaler determination circuit being responsive to and combining the recursive vector determination circuit output signal and a delayed covariance matrix signal and providing a recursive scaler determination circuit output signal in response thereto;

a recursive gain determination circuit operatively coupled to the recursive scaler determination circuit, the recursive gain determination circuit being responsive to the recursive scaler determination circuit output signal and providing a recursive gain determination circuit output signal corresponding to the processing circuit output signal in response thereto;

a covariance matrix determination circuit operatively coupled to the recursive gain determination circuit, the covariance matrix determination circuit being responsive to the recursive gain determination circuit output signal and providing a covariance matrix determination circuit output signal in response thereto; and

a delay circuit operatively coupled to the covariance matrix determination circuit and being responsive to the covariance matrix determination circuit output signal, the delay circuit also being operatively coupled to the recursive scaler determination circuit, the delay circuit effectively providing the delayed covariance matrix signal to the recursive scaler determination circuit.

7. A target prediction and collision warning system as defined by claim 6, wherein the recursive scaler determination circuit comprises:

a transpose circuit operatively coupled to recursive vector determination circuit, the transpose circuit being responsive to the recursive vector determination circuit output signal and generating a vector transpose circuit output signal,

a scaler mixer circuit operatively coupled to the transpose circuit and being responsive to the vector transpose circuit output signal, the scaler mixer circuit also being operatively coupled to the recursive vector determination circuit and being responsive to the recursive vector determination circuit output signal, the scaler mixer circuit also being coupled to the delay circuit and being responsive to the delayed covariance matrix signal, the scaler mixer circuit effectively combining the vector transpose circuit output signal, the recursive vector determination circuit output signal and the delayed covariance matrix signal and providing a scaler mixer circuit output signal in response thereto,

a scaler combiner circuit operatively coupled to the scaler mixer circuit and being responsive to the scaler mixer circuit output signal, the scaler combiner circuit also being responsive to a second signal as an input signal, the scaler combiner circuit effectively combining the scaler mixer circuit output signal and the second signal and providing a scaler combiner circuit output signal in response thereto; and

an inverse determination circuit coupled to the scaler combiner circuit, the inverse determination circuit being responsive to the scaler combiner circuit output signal and providing an inverse determination circuit output signal in response thereto corresponding to the recursive scaler determination circuit output signal.

8. A target prediction and collision warning system as defined by claim 7, wherein the recursive gain determination circuit comprises:

a recursive gain mixer circuit operatively coupled to the transpose circuit and being responsive to the vector transpose circuit output signal, the recursive gain mixer circuit also being operatively coupled to the inverse determination circuit and being responsive to the recursive scaler determination circuit output signal, the recursive gain mixer circuit also being operatively coupled to the delay circuit and being responsive to the delayed covariance matrix signal, the gain mixer circuit effectively mixing the vector transpose circuit output signal, the inverse determination circuit output signal and the delayed covariance matrix signal and providing a gain mixer circuit output signal indicative of a signal gain factor in response thereto.

9. A target prediction and collision warning system as defined by claim 6, wherein the covariance matrix determination circuit comprises:

an output port;

a first covariance mixer circuit operatively coupled to the gain mixer circuit and being responsive to the gain mixer circuit output signal, the first covariance mixer circuit also being operatively coupled to the transpose circuit and being responsive to the vector transpose circuit output signal, the first covariance mixer circuit effectively mixing the gain mixer circuit output signal and the transpose circuit output signal and providing a first covariance mixer circuit output signal in response thereto;

a covariance combiner circuit operatively coupled to the first covariance mixer circuit and being responsive to the first covariance mixer circuit output signal, the covariance combiner circuit also being responsive to the second signal as an input signal, the covariance combiner circuit effectively combining the first covariance mixer circuit output signal and the second signal and providing a covariance combiner circuit output signal in response thereto; and

a second covariance mixer circuit operatively coupled to the covariance combiner circuit and being responsive to the covariance combiner circuit output signal, the second covariance mixer circuit also being operatively coupled to the delay circuit and being responsive to the delayed covariance matrix signal, the covariance matrix determination circuit effectively combining covariance combiner circuit output signal and the delayed covariance matrix signal and generating a covariance matrix determination output signal in response thereto.

10. A target prediction and collision warning system as defined by claim 4, further comprising a data recorder operatively coupled to the second recursive combiner circuit, wherein the output signal of the Kalman filter circuit comprises regression coefficients describing the relationship of relative range and velocity of at least one target in the spatial region, the data recorder receiving and storing the coefficients for future evaluation.

11. A target prediction and collision warning system for detecting targets in a spatial region about a vehicle comprising:

a) a signal transmitter providing at least first and second detection signals to the spatial region, the at least first and second detection signals at least partially reflecting off at least one object within the spatial region thereby generating at least partially reflected first and second detection signals, each of the at least partially reflected first and second detection signals being indicative of at least one characteristic of the at least one object within the spatial region;

(b) a signal receiver operatively coupled to the signal transmitter, the signal receiver receiving each of the at least partially reflected first and second detection signals, each of the at least partially reflected first and second detection signals corresponding to the first and second object parameter signals;

c) a target tracker circuit operatively coupled to the signal receiver, the target tracker circuit comprises a direct least squares determination circuit, the direct least squares determination circuit further comprising;

a direct vector determination circuit operatively coupled to the signal receiver, the direct vector determination circuit being responsive to at least one of the second object parameter signals and generating a direct vector determination circuit output signal in response thereto;

a matrix determination circuit operatively coupled to the signal receiver and being responsive to at least one of the first object parameter signals and generating a matrix determination output signal in response thereto;

a matrix transposition circuit operatively coupled to the matrix determination circuit and receiving the matrix determination circuit output signal and generating a matrix transposition circuit output signal in response thereto corresponding to a transpose function of the matrix determination circuit output signal;

a first direct mixer circuit operatively coupled to the matrix determination circuit and the matrix transposition circuit, the first direct mixer circuit effectively combining the matrix determination circuit output signal and the matrix transposition circuit output signal and providing a first direct mixer output signal in response thereto;

a second direct mixer circuit operatively coupled to the direct vector determination circuit and the matrix transposition circuit, the second direct mixer circuit being responsive to and combining the direct vector determination circuit output signal and the matrix transposition circuit output signal and providing a second direct mixer circuit output signal in response thereto;

an inverse determination circuit operatively coupled to the first direct mixer circuit and being responsive to the first direct mixer circuit output signal, the inverse determination circuit generating and providing an output signal corresponding to an inverse matrix function of the first direct mixer circuit output signal in response thereto;

a third direct mixer circuit operatively coupled to the inverse determination circuit and the second direct mixer circuit, the third direct mixer circuit being responsive to and combining the second direct mixer circuit output signal and the inverse determination circuit output signal and providing a third direct mixer circuit output signal in response thereto, the third direct mixer circuit output signal corresponding to a direct least squares determination circuit signal, the least squares determination circuit being responsive to at least one of the first and second object parameter signals and providing a regression parameter signal indicative of at least one of a past, current and future value of the at least one characteristic of the at least one object within the spatial region, whereby the present and future position of a target within a spatial region may be predicted and imminent collision warning indicated.

12. A target prediction and collision warning system as defined by claim 11, wherein the target tracker further comprises a recursive least squares determination circuit coupled to the direct least squares determination circuit, the recursive least squares determination circuit being responsive to the direct least squares determination circuit signal and the first and second object parameter signals and providing a current regression parameter signal.

13. A target prediction and collision warning system as defined by claim 12, wherein the recursive least squares determination circuit comprises:

a recursive vector determination circuit operatively coupled to the signal receiver, the recursive vector determination circuit being responsive to the second object parameter signal and generating a recursive vector determination circuit output signal in response thereto;

a first recursive mixer circuit operatively coupled to the recursive vector determination circuit, the first recursive mixer circuit being responsive to the recursive vector determination circuit output signal and regression parameter signal, the first recursive mixer circuit combining the recursive vector determination circuit output signal and the initial regression parameter signal and providing a first recursive mixer circuit output signal in response thereto;

a first recursive combiner circuit operatively coupled to the first recursive mixer circuit and being responsive to the first recursive mixer circuit output signal, the first recursive combiner circuit also being operatively coupled to the signal receiver and being responsive to the first object parameter signal, the first recursive combiner circuit effectively combining the first recursive mixer circuit output signal and the first object parameter signal and providing a first recursive combiner circuit output signal in response thereto;

a processing circuit operatively coupled to the recursive vector determination circuit and being responsive to the recursive vector determination circuit output signal and providing a processing circuit output signal in response thereto;

a second recursive mixer circuit operatively coupled to the first recursive combiner circuit and the processing circuit, the second recursive mixer circuit being responsive to and combining the first recursive combiner circuit output signal and the processing circuit output signal, the second mixer circuit providing an output signal in response thereto; and

a second recursive combiner circuit operatively coupled to the second recursive mixer circuit, the second recursive combiner circuit having an output port corresponding to an output port of the recursive least squares determination circuit, the second recursive combiner circuit being responsive to and combining the second recursive mixer circuit output signal and the regression parameter signal and providing an output signal indicative of an updated regression parameter signal.

14. A target prediction and collision warning system as defined by claim 13, wherein the recursive least squares determination circuit further comprises:

a delay circuit operatively coupled to the second recursive combiner circuit output port and being responsive to the updated regression parameter signal and generating the regression parameter signal, the delay circuit also being operatively coupled to the first recursive mixer circuit, the delay circuit providing the regression parameter signal to the first recursive mixer circuit and the second recursive combiner circuit.

15. A target prediction and collision warning system as defined by claim 13, wherein the processing circuit comprises:

a recursive scaler determination circuit operatively coupled to the recursive vector determination circuit, the recursive scaler determination circuit being responsive to and combining the recursive vector determination circuit output signal and a delayed covariance matrix signal and providing a recursive scaler determination circuit output signal in response thereto;

a recursive gain determination circuit operatively coupled to the recursive scaler determination circuit, the recursive gain determination circuit being responsive to the recursive scaler determination circuit output signal and providing a recursive gain determination circuit output signal in response thereto corresponding to the processing circuit output signal in response thereto;

a covariance matrix determination circuit operatively coupled to the recursive gain determination circuit, the covariance matrix determination circuit being responsive to the recursive gain determination circuit output signal and providing a covariance matrix determination circuit output signal in response thereto; and

a delay circuit operatively coupled to the covariance matrix determination circuit and being responsive to the covariance matrix determination circuit output signal, the delay circuit also being operatively coupled to the recursive scaler determination circuit, the delay circuit providing the delayed covariance matrix signal to the recursive scaler determination circuit.

16. A target prediction and collision warning system as defined by claim 15, further comprising:

a switching circuit operatively coupled to the direct least square determination circuit and to the covariance matrix determination circuit, the switching circuit also being coupled to at least one of the recursive scaler determination circuit, the gain determination circuit and the covariance matrix determination circuit, the switching circuit selectively coupling the at least one of the recursive scaler determination circuit, the gain determination circuit and the covariance matrix determination circuit to one of the direct least squares determination circuit and the covariance matrix determination circuit.

17. A method of detecting and tracking motion of objects in a spatial region proximate to a vehicle for collision avoidance, the method comprising the steps of:

a) transmitting at least first and second detection signals for at least partial reflection by at least one object in the spatial region thereby generating at least partially reflected first and second detection signals;

b) detecting the at least partially reflected first and second detection signals, the at least partially reflected first and second detection signals corresponding to first and second object parameter signals;

c) combining the first and second parameter signals to generate a regression parameter signal indicative of at least one of a post, current and future characteristic of the at least one object by the further steps of;

1) obtaining at least three sample value signals for each of the first and second object parameter signals;

2) generating a direct least squares signal utilizing the at least three sample value signals for each of the first and second object parameter signals, the direct least squares signal corresponding to an initial estimate of the regression parameter signal;

3) obtaining at least a fourth sample value signal for each of the first and second object parameter signals;

4) generating a recursive least squares signal utilizing the direct least squares signal, the at least three sample value signals and the at least fourth sample value signal for each of the first and second object parameter signals, the recursive least squares signal corresponding to the regression parameter signal, the generation of the recursive least squares signal further comprising the steps of;

A) generating a first object parameter vector signal utilizing the at least three sample value signals and the at least fourth sample value signal for the first object parameter signal;

B) mixing the first object parameter vector signal and the direct least squares signal to provide a first mixed signal;

C) generating a second object parameter vector signal utilizing the at least three sample value signals and the at least fourth sample value signal for the first object parameter signal;

D) combining the first mixed signal and the second object parameter vector signal;

E) generating a difference signal indicative of a difference between the at least three sample value signals of the first and second object parameter signals and the at least fourth sample value signal of the first and second object parameter signals;

F) mixing the difference signal and the first combiner circuit output signal to provide a second mixed signal; and

G) combining the second mixed signal and the direct least squares signal to provide the recursive least squares signal, the recursive least squares signal indicative of at least one of the present and future position of a target within a spatial region whereby imminent collision warning may be indicated to facilitate collision avoidance.

18. A method as defined by claim 17, which further comprises:

generating an updated recursive least squares signal based upon the at least three sample value signals, the at least fourth sample value signals and at least a fifth sample value signal for the at least first and second object parameter signals.

19. A method as defined by claim 18, further comprising:

generating an updated second object parameter signal based upon the at least three sample value signals, the at least fourth sample value signals and at least a fifth sample value signal for the at least second parameter signals;

mixing the updated second object parameter signal with the recursive least squares signal to provide an updated first mixed signal;

generating an updated first object parameter signal utilizing the at least three sample value signals, the at least fourth sample value signal and a fifth sample value signal for the at least first parameter signals;

combining the updated first mixed signal and the updated first object parameter signal to provide an updated first combined signal;

generating an updated signal indicative of difference between the at least three and fourth sample value signals of the first and second object parameter signals and the at least fifth sample value signals of the first and second object parameter signals;

mixing the updated signal and the updated first combiner circuit output signal to provide an updated second mixed signal; and

combining the updated second mixed signal and the recursive least squares signal to provide an updated second combined signal corresponding to an updated recursive least squares signal.

20. A method as defined by claim 17, wherein step (2) further comprises:

generating a second object parameter vector signal utilizing the at least three sample value signals for the second object parameter signal;

generating a first object parameter matrix signal utilizing the at least three sample value signals for the first object parameter signal;

generating a first object parameter transposition matrix signal utilizing the first object parameter matrix signal indicative of a transpose matrix function of the first object parameter matrix signal;

combining the first object parameter matrix signal and the first object parameter transposition matrix signal to provide a first mixed signal;

combining the first object parameter transposition matrix signal and the first object parameter matrix signal to provide a second mixed signal;

generating an inverse matrix signal utilizing the first mixed signal;

combining the inverse matrix signal and the second mixed signal to provide a third mixed signal corresponding to the direct least square signal.

21. A method as defined by claim 17, wherein the recursive least squares signal comprises regression coefficients describing the relationship of the relative range and velocity of a target in a spatial region and the method further includes the step of recording the coefficients for future evaluation.