Systems and methods involving features of adaptive and/or autonomous traffic control

• US 8,825,350 B1
• Filed: 02/08/2012
• Issued: 09/02/2014
• Est. Priority Date: 11/22/2011
• Status: Active Grant

• Current Assignee: Zongzhi Li
• Original Assignee: Kurt B. Robinson
• Inventors: Robinson, Kurt B.
• Primary Examiner(s): Khatib, Rami
• Application Number:
  US13/369,233
• Time in Patent Office: 937 Days
• Field of Search: 701/1, 701/116, 701/117, 701/118, 701/119, 340/933, 340/934, 340/935, 340/936, 340/907, 340909-915, 340/919
• US Class Current: 701/117
• CPC Class Codes: G06V 20/54 of traffic, e.g. cars on the road, trains or boats
• CPC Class Codes: G08G 1/0116 from roadside infrastructure, e.g. beacons
• CPC Class Codes: G08G 1/0145 for active traffic flow control
• CPC Class Codes: G08G 1/042 using inductive or magnetic detectors

Abstract

Systems and method are disclosed for adaptive and/or autonomous traffic control. In one illustrative implementation, there is provided a method for processing traffic information. Moreover, the method may include receiving data regarding travel of vehicles associated with an intersection, using neural network technology to recognize types and/or states of traffic, and using the neural network technology to process/determine/memorize optimal traffic flow decisions as a function of experience information. Exemplary implementations may also include using the neural network technology to achieve efficient traffic flow via recognition of the optimal traffic flow decisions.

Claims

1. A method for processing traffic information, the method comprising:

receiving data regarding travel of vehicles associated with an intersection;

using neural network technology to recognize traffic type, traffic types being composed of one or both of traffic context types and traffic category types;

using the neural network technology to recognize states of the traffic;

using the neural network technology to memorize optimal traffic flow decisions as a function of prior experience;

using the neural network technology to achieve efficient traffic flow via recognition and use of the optimal traffic flow decisions in providing an output that controls a traffic light system; and

enhancing recognition of the types of traffic using an array having a plurality of sensor inputs beyond inputs from video sensors and physical infrastructure;

wherein the other sensor inputs include an infrared traffic input providing enhanced capability from transmission of various data types;

wherein the infrared traffic input comprises;

detailed infrared signatures of detected traffic objects independent from adverse weather or light conditions for recognition of vehicle position and type; and

detection of vehicle passenger occupancy which is used in higher level traffic flow decision making;

and wherein the method further comprises;

giving priority to passenger-weighted incoming traffic.

2. The method of claim 1 further comprising:

processing new sensory inputs;

retraining previously deployed neurons to take advantage of the new sensors;

wherein recognition of the traffic types is enhanced via adapting their recognition to utilize the new inputs.

3. The method of claim 1 further comprising:

processing new sensory inputs;

utilizing additional, reserve neuron capacity that was initially included in the system, wherein the reserve neuron capacity is subsequently trained to utilize the new inputs.

4. The method of claim 1 further comprising:

processing the data from expanded sensory input(s) to detect small traffic objects defined as traffic objects that are smaller than cars, wherein the system is configured to adapt to the small traffic objects; and

incorporating the small traffic object information into the higher level traffic flow control decision-making.

5. The method of claim 1 further comprising:

utilizing 2 or more layers of neural network neuron storage elements for unique recognition, classification, and/or traffic flow decision-making tasks.

6. The method of claim 5 wherein the unique recognition includes feeding from a low level layer to higher level layers.

7. The method of claim 5 wherein the traffic flow decision-making includes state information inputs from nearby similar traffic controllers comparable with intermediate level traffic state recognition at the local controller which are combined at higher levels to optimize traffic flow decision-making at the local traffic signal light.

8. The method of claim 1 further comprising enhancing the traffic type recognition using an array having a plurality of sensor inputs beyond inputs from video sensors and traditional physical infrastructure.

9. The method of claim 8 wherein the other sensor inputs include an infrared traffic input providing enhanced capability from transmission of various data types.

10. The method of claim 9 wherein the infrared traffic input comprises:

detailed infrared signatures of detected traffic objects independent from adverse weather or light conditions for recognition of vehicle position and type; and

detection of vehicle passenger occupancy which is used in higher level traffic flow decision making;

and wherein the method further comprises;

giving priority to passenger-weighted incoming traffic.

11. A method for processing traffic information, the method comprising:

receiving data regarding travel of vehicles associated with an intersection;

using neural network technology to recognize traffic type, traffic types being composed of one or both of traffic context types and traffic category types;

using the neural network technology to recognize states of the traffic;

using the neural network technology to memorize optimal traffic flow decisions as a function of prior experience; and

using the neural network technology to achieve efficient traffic flow via recognition and use of the optimal traffic flow decisions in providing an output that controls a traffic light system;

wherein higher levels of neural net recognition hierarchy are used to determine relative traffic object and pedestrian positions and speeds, which is used to select optimal traffic light sequences, eliminating digital processing delays and associated costs using a sequence of neural network recognition cycles.

12. The method of claim 11 further comprising enhancing the traffic type recognition using an array having a plurality of sensor inputs beyond inputs from video sensors and traditional physical infrastructure.

13. The method of claim 12 wherein the other sensor inputs include an infrared traffic input providing enhanced capability from transmission of various data types.

14. The method of claim 13 wherein the infrared traffic input comprises:

detailed infrared signatures of detected traffic objects independent from adverse weather or light conditions for recognition of vehicle position and type; and

detection of vehicle passenger occupancy which is used in higher level traffic flow decision making;

and wherein the method further comprises;

giving priority to passenger-weighted incoming traffic.

15. The method of claim 12 further comprising:

processing new sensory inputs;

retraining previously deployed neurons to take advantage of the new sensors;

wherein recognition of the traffic types is enhanced via adapting their recognition to utilize the new inputs.

16. The method of claim 12 further comprising:

processing new sensory inputs;

utilizing additional, reserve neuron capacity that was initially included in the system, wherein the reserve neuron capacity is subsequently trained to utilize the new inputs.

17. The method of claim 12 further comprising:

processing the data from expanded sensory input(s) to detect reduced traffic flow, wherein the system is configured to adapt to the reduced traffic flow; and

incorporating the reduced traffic flow information into the higher level traffic flow control decision-making.

18. The method of claim 12 further comprising:

utilizing 2 or more layers of neural network neuron storage elements for unique recognition, classification, and/or traffic flow decision-making tasks.

19. The method of claim 18 wherein the unique recognition includes feeding from a low level layer to higher level layers.

20. The method of claim 18 wherein the traffic flow decision-making includes state information inputs from nearby similar traffic controllers comparable with intermediate level traffic state recognition at the local controller which are combined at higher levels to optimize traffic flow decision-making at the local traffic signal light.

21. The method of claim 11 further comprising:

processing new sensory inputs;

retraining previously deployed neurons to take advantage of the new sensors;

wherein recognition of the traffic types is enhanced via adapting their recognition to utilize the new inputs.

22. The method of claim 11 further comprising:

processing new sensory inputs;

utilizing additional, reserve neuron capacity that was initially included in the system, wherein the reserve neuron capacity is subsequently trained to utilize the new inputs.

23. The method of claim 11 further comprising:

processing the data from expanded sensory input(s) to detect reduced traffic flow, wherein the system is configured to adapt to the reduced traffic flow; and

incorporating the reduced traffic flow information into the higher level traffic flow control decision-making.

24. The method of claim 11 further comprising:

utilizing 2 or more layers of neural network neuron storage elements for unique recognition, classification, and/or traffic flow decision-making tasks.

25. The method of claim 24 wherein the unique recognition includes feeding from a low level layer to higher level layers.

26. The method of claim 24 wherein the traffic flow decision-making includes state information inputs from nearby similar traffic controllers comparable with intermediate level traffic state recognition at the local controller which are combined at higher levels to optimize traffic flow decision-making at the local traffic signal light.