Traffic surveillance and simulation apparatus

• US 5,801,943 A
• Filed: 03/06/1995
• Issued: 09/01/1998
• Est. Priority Date: 07/23/1993
• Status: Expired due to Fees

• Current Assignee: Condition Monitoring Systems Incorporated
• Original Assignee: Condition Monitoring Systems Incorporated
• Inventors: Nasburg, Robert E.
• Primary Examiner(s): Trans, Vincent N.
• Application Number:
  US08/398,770
• Time in Patent Office: 1,275 Days
• Field of Search: 364/436, 364/439, 340/905, 340/910
• US Class Current: 701/117
• CPC Class Codes: G08G 1/0104 Measuring and analyzing of parameters relative to traffic conditions
• CPC Class Codes: G08G 1/07 Controlling traffic signals
• CPC Class Codes: G08G 1/096716 where the received information does not generate an automatic action on the vehicle control
• CPC Class Codes: G08G 1/096741 where the source of the transmitted information selects which information to transmit to each vehicle
• CPC Class Codes: G08G 1/096775 where the origin of the information is a central station
• CPC Class Codes: G08G 1/20 Monitoring the location of vehicles belonging to a group, e.g. fleet of vehicles, countable or determined number of vehicles

Abstract

A wide area surveillance system for application to large road networks is described. The system employs smart sensors to identify plural individual vehicles in the network. These vehicles are tracked on an individual basis, and the system derives the behavior of the vehicle. Furthermore, the system derives traffic behavior on a local basis, across roadway links, and in sections of the network. Processing in the system is divided into multiple processing layers, with geographical separation of tasks.

Claims

1. A traffic surveillance system comprising:

a first sensor positioned to sense vehicular traffic in a predetermined first field;

a first local traffic processor coupled to the first sensor for identifying vehicles within the field of the first sensor by periodically sampling the first sensor and extracting vehicle locations and identification information;

a second sensor positioned to sense vehicular traffic in a second predetermined field where said second predetermined field is separated from said first predetermined field;

a second local traffic processor coupled to the second sensor for identifying vehicles within the field of the second sensor by periodically sampling the second sensor and extracting vehicle locations and identification information; and

a wide area traffic flow processor coupled to each local traffic processor for receiving the vehicle location and identification information from each local traffic processor and tracking the identified vehicles;

wherein the wide area traffic flow processor utilizes a predictor algorithm to predict each identified vehicle'"'"'s location, and utilizes the vehicle location and identification information to correct the predictor model thereby functioning to monitor traffic consisting of all the identified vehicles.

2. A traffic surveillance system as set forth in claim 1, wherein the wide area traffic flow processor predicts each vehicle'"'"'s location at approximately the same time as the local traffic processor is sampling the sensor.

3. A traffic surveillance system comprising:

a first sensor positioned to sense vehicular traffic in a predetermined first field;

a first local traffic processor for identifying vehicles within the first field and periodically generating vehicle locations and identification information concerning the vehicles in the first field;

a second sensor Dositioned to sense vehicular traffic in a predetermined second field;

a second local traffic processor for identifying vehicles within the second field and periodically generating vehicle locations and identification information concerning the vehicles in the second field; and

a wide area traffic flow processor coupled to the first and second local traffic processors for receiving the vehicle location and identification information from the local traffic processors and tracking the identified vehicles, the wide area traffic flow processor reporting predictions of where the vehicles will be at the next period;

wherein the wide area traffic flow processor models the kinematics of the vehicles individually, and utilizes new vehicle location and identification information to correct the prediction mechanism.

4. A traffic surveillance system as set forth in claim 3, wherein the wide area traffic flow processor generates a micromodel of vehicle behavior, the micromodel being a GM car following micromodel.

5. A traffic surveillance system as set forth in claim 4, wherein the predictor algorithm further incorporates lane gap calculations, the lane gap calculations modelling the influence of roadway geometric events.

6. A traffic surveillance system as set forth in claim 5, wherein the roadway geometric events include movement from a straight lane segment to a curved lane segment.

7. A traffic surveillance system as set forth in claim 5, wherein the roadway geometric events include movement from a curved lane segment to a straight lane segment.

8. A traffic surveillance system as set forth in claim 5, wherein the roadway geometric events include changing traffic control signal conditions.

9. A traffic surveillance system as set forth in claim 3, wherein the roadway geometric events include requirements to stop at defined locations.

10. A traffic surveillance system as set forth in claim 3, wherein the roadway geometric events include availability of a lane into which an identified vehicle may merge.

11. A traffic surveillance system as set forth in claim 3, wherein the roadway geometric events include availability of a new lane segment into which an identified vehicle may move.

12. An apparatus for tracking vehicles in a roadway comprising a programmed computer system, the roadway including first and second lanes, wherein the first lane includes a segment which is substantially parallel to a segment of the second lane such that a vehicle could move directly therebetween, the beginning of the substantially parallel segments defining a hypothesis generation point, and the end of the substantially parallel segments defining a decision hold off point in each lane, there being further defined a first decision point in the first lane beyond the decision hold off point and a second decision point in the second lane beyond the decision hold off point, the program being selectively operable to effect the process of:

obtaining a vehicle'"'"'s location on the first lane segment within a predetermined distance of the hypothesis generation point in a first period;

generating a hypothetical track corresponding to the vehicle continuing on the first lane and a hypothetical track corresponding to the vehicle moving to the second lane;

generating a probability of correctness for the hypotheses;

obtaining the vehicle'"'"'s location in a second period after the first period;

updating the probabilities based upon the vehicle'"'"'s location in the second period;

determining if the vehicle'"'"'s location is past the decision hold off point, and if so, then determining whether one of the probabilities has reached a threshold, and if so, then eliminating the track associated with the hypothesis which did not reach the threshold, and otherwise updating the tracks with the vehicle'"'"'s location in the second period and continuing with the method at the updating probabilities step.

13. A programmed computer system as set forth in claim 12, further comprising a display, wherein the program includes instructions for displaying representations of the tracked vehicles on the display.

14. A programmed computer system as set forth in claim 13, wherein a simulation of the roadway is displayed on the display, and the representations of the tracked vehicles are overlaid on the displayed roadway simulation.

15. A method of tracking vehicles comprising:

obtaining current vehicle location and generating a vehicle track from the location of the vehicle;

extending the vehicle track based upon the current location of the vehicle;

associating and updating all vehicle tracks;

for each vehicle track;

performing a first test of whether the vehicle track has reached a hypothesis generation point, and if not then continuing with the method at the vehicle location obtaining step, and otherwise;

extending all vehicle tracks;

performing associations between vehicle tracks;

updating all vehicle tracks;

updating hypothesis probabilities for the vehicle tracks;

updating vehicle track types;

testing if the vehicle tracks have passed a hold off point, and if not then continuing with the method at the extending all vehicle tracks step, and otherwise

testing if a stopping rule has been satisfied, and if not then continuing with the method at the extending all vehicle tracks step, and otherwise

continuing with the method at the first test step.

16. A method as set forth in claim 15, further comprising the step of display icons representative of the tracked vehicles on a display coupled to the computer.

17. A method of tracking a first vehicle on a roadway, the roadway comprising a first lane segment contiguous with a second lane segment, and a plurality of geometric events having predetermined locations, the method comprising:

(a) obtaining the first vehicle'"'"'s location on the first lane segment in a first time period, and the first vehicle'"'"'s location on the first lane segment in a second time period;

(b) predicting a track of the first vehicle as an extension of the first vehicle'"'"'s locations in the first and second time periods;

(c) determining if geometric events are in the first vehicle'"'"'s track, and if present, identifying the geometric events;

(d) determining if a second vehicle is preceding the first vehicle, and, if present, obtaining the second vehicle'"'"'s kinematics;

(e) determining the lane gap, the car gap, the lane influence point and the headway influence point for the first vehicle;

(f) if the lane gap is greater than the lane influence point and the car gap is greater than the headway influence point, then extending the first vehicle'"'"'s track without taking into account the geometric events or the second vehicle'"'"'s kinematics;

(g) if the lane gap is greater than the lane influence point and the car gap is not greater than the headway influence point, then extending the first vehicle'"'"'s track without taking into account the geometric events while taking into account the second vehicle'"'"'s kinematics;

(h) if the lane gap is not greater than the lane influence point and the car gap is greater than the headway influence point, then extending the first vehicle'"'"'s track taking into account the geometric events and not taking into account the second vehicle'"'"'s kinematics;

(i) if the lane gap is not greater than the lane influence point and the car gap is not greater than the headway influence point, then extending the first vehicle'"'"'s track taking into account the geometric events and the second vehicle'"'"'s kinematics.

18. A method of tracking a first vehicle on a roadway as set forth in claim 17, further comprising displaying the first vehicle'"'"'s predicted location on a display.

19. A method of tracking a first vehicle on a roadway as set forth in claim 17, wherein a nominal velocity is associated with the first lane segment, and if the lane gap is greater than the lane influence point and the car gap is greater than the headway influence point, then predicting that the first vehicle'"'"'s velocity will approach the nominal velocity.

20. A method of tracking a first vehicle on a roadway as set forth in claim 17, wherein if the lane gap is greater than the lane influence point and the car gap is not greater than the headway influence point, then predicting the first vehicle'"'"'s kinematics from the GM micromodel.

21. A method of tracking a first vehicle on a roadway as set forth in claim 17, wherein the second lane segment is a continuation of the first lane segment and a nominal velocity is associated with the second lane segment, and if the lane gap is not greater than the lane influence point and the car gap is greater than the headway influence point, then predicting that the first vehicle'"'"'s kinematics will approach the nominal velocity.

22. A method of tracking a first vehicle on a roadway as set forth in claim 17, wherein if the lane gap is not greater than the lane influence point and the car gap is not greater than the headway influence point, then predicting that the first vehicle'"'"'s will accelerate at the lesser of the acceleration from the first vehicle'"'"'s velocity in the first period necessary to approach the nominal velocity of the second lane segment and the acceleration from the first vehicle'"'"'s velocity in the first period to approach the velocity of the second vehicle.

23. A method of tracking a first vehicle on a roadway as set forth in claim 17, wherein the second lane segment begins at a location noncontiguous of the first lane segment, the first lane segment ends at the second lane segment and the second lane segment continues, and the geometric events include a yield segment located at the junction of the first and second lane segments, the method further comprising, in the extending step, if the predicted track of the first vehicle includes the yield segment then moving the first vehicle into the second segment when there is sufficient space.

24. A method of tracking a first vehicle on a roadway as set forth in claim 23, wherein a nominal velocity is associated with the second lane segment, and if a second vehicle on the second lane segment is predicted to be within a predetermined spacing distance of the yield segment at the same time the first vehicle is predicted to arrive at the yield segment, then the velocity associated with the yield segment is set to zero, and otherwise the velocity of the yield segment is set to the nominal velocity of the second lane segment.

25. A method of tracking a first vehicle on a roadway as set forth in claim 17, wherein the first lane segment is parallel to the second lane segment and the first lane segment ends at a location where the second lane segment continues, and the geometric events include a merge segment continuing from the first lane segment and ending at a location along the second lane segment, and if the predicted track of the first vehicle includes the merge segment, then the method further comprising:

waiting a number of periods of time until the track of the first vehicle is predicted to be in the merge segment;

determining the sufficiency of intervehicle spacing in the second lane segment along and behind the first vehicle;

selecting the location of closest sufficient spacing on the second lane segment for the first vehicle to merge into the second lane segment;

determining the lane gap to be equal to the distance between the first vehicle and the middle of the closest sufficient spacing;

waiting a number of periods of time until the first vehicle is parallel to the location on the second lane segment of the sufficient spacing, and predicting the first vehicle to merge into second lane at the same velocity as the first vehicle had in the merge segment.

26. A method of tracking a first vehicle on a roadway as set forth in claim 17, wherein the first lane segment ends at the beginning of the second lane segment and a signal is located at the junction of the first and second lane segments, a nominal velocity is associated with the second lane segment, and the geometric events include a signalled stop bar segment located at the junction of the first and second lane segments, and if the predicted track of the first vehicle includes the signalled stop bar segment then determining the state of the signal, and if the signal phase indicates a stopping condition then setting the velocity of the signalled stop bar segment to be zero, and if the signal phase indicates a proceed condition then setting the velocity of the signalled stop bar to be the nominal velocity of the second lane segment.

27. A method of tracking a first vehicle on a roadway as set forth in claim 17, wherein the second lane begins at a location noncontiguous of the first lane segment, the first lane segment ends at the second lane segment and the second lane segment continues, the first lane segment is curved and the second lane segment is straight, the geometric events include a transition segment located at the junction of the first and second lane segments, and if the predicted track of the first vehicle includes the transition segment then predicting that the first vehicle will maintain its velocity through the first lane segment and will begin to accelerate after passing the transition segment.

28. A traffic surveillance system comprising:

plural local traffic processors each for identifying vehicles within a field and periodically generating vehicle locations and identification information, the fields including first and second lanes having at least one segment each which are substantially parallel such that a vehicle could move directly therebetween, the beginning of the substantially parallel segments defining a hypothesis generation point, and the end of the substantially parallel segments defining a decision hold off point in each lane, there being further defined a first decision point in the first lane beyond the decision hold off point and a second decision point in the second lane beyond the decision hold off point; and

a wide area traffic flow processor coupled to the local traffic processors for receiving the vehicle location and identification information from the local traffic processors and tracking the identified vehicles, the wide area traffic flow processor reporting predictions of where the vehicles will be at the next period, and being selectively operable to effect the process of;

obtaining a vehicle'"'"'s location on the first lane segment within a predetermined distance of the hypothesis generation point in a first period;

generating a hypothetical track corresponding to the vehicle continuing on the first lane and a hypothetical track corresponding to the vehicle moving to the second lane;

generating a probability of correctness for the hypotheses;

obtaining the vehicle'"'"'s location in a second period after the first period;

updating the probabilities based upon the vehicle'"'"'s location in the second period;

determining if the vehicle'"'"'s location is past the decision hold off point, and if so, then determining whether one of the probabilities has reached a threshold, and if so, then eliminating the hypothetical track associated with the hypothesis which did not reach the threshold, and otherwise updating the hypothetical tracks with the vehicle'"'"'s location in the second period and continuing with the method at the updating probabilities step;

wherein the wide area traffic flow processor models the kinematics of the vehicles individually, and utilizes new vehicle location and identification information to correct the prediction mechanism.

29. A traffic surveillance system comprising:

plural local traffic processors each for identifying vehicles within a field and periodically generating vehicle locations and identification information, the fields including first and second contiguous lane segments and plural geometric events; and

a wide area traffic flow processor coupled to the local traffic processor for receiving the vehicle location and identification information from the local traffic processor and tracking the identified vehicles, the wide area traffic flow processor reporting predictions of where the vehicles will be at the next period and being selectively operable to effect the process of;

obtaining a first vehicle'"'"'s location on the first lane segment in a first time period, and the first vehicle'"'"'s location on the first lane segment in a second time period;

predicting a track of the first vehicle as an extension of the first vehicle'"'"'s locations in the first and second time periods;

determining if geometric events are in the first vehicle'"'"'s track, and if present, identifying the geometric events;

determining if a second vehicle is preceding the first vehicle, and, if present, obtaining the second vehicle'"'"'s kinematics;

determining the lane gap, the car gap, the lane influence point and the headway influence point for the first vehicle;

if the lane gap is greater than the lane influence point and the car gap is greater than the headway influence point, then extending the first vehicle'"'"'s track without taking into account the geometric events or the second vehicle'"'"'s kinematics;

if the lane gap is greater than the lane influence point and the car gap is not greater than the headway influence point, then extending the first vehicle'"'"'s track without taking into account the geometric events and taking into account the second vehicle'"'"'s kinematics;

if the lane gap is not greater than the lane influence point and the car gap is greater than the headway influence point, then extending the first vehicle'"'"'s track taking into account the geometric events and not taking into account the second vehicle'"'"'s kinematics;

if the lane gap is not greater than the lane influence point and the car gap is not greater than the headway influence point, then extending the first vehicle'"'"'s track taking into account the geometric events and the second vehicle'"'"'s kinematics;

wherein the wide area traffic flow processor models the kinematics of the vehicles individually, and utilizes new vehicle location and identification information to correct the prediction mechanism.