Vehicle path planning
A computing system can be programmed to determine a location, speed and direction for a first vehicle. The computer can be further programmed to determine probabilities for predicted locations, speeds and directions of the first vehicle based on identifying the first vehicle and the driving environment. The computer can be further programmed to operate a second vehicle based on the determined probabilities for predicted locations, speeds and directions of the first vehicle.
1. A method, comprising:
determining a plurality of possible first vehicle path polynomials including predicted locations, speeds and directions of a first vehicle, and further determining respective probabilities of each of the path polynomials, based on (a) identifying the first vehicle according to at least one of a type of the first vehicle or a human driver of the first vehicle, and (b) a driving environment in a second vehicle;
determining a second vehicle planned path polynomial for the second vehicle based on the determined probabilities corresponding to the predicted first vehicle path polynomials; and
operating the second vehicle based on the second vehicle planned path polynomial.
2. The method of claim 1 , further comprising predicting the location, speed and direction for the first vehicle based on vehicle sensor data.
3. The method of claim 2 , wherein processing vehicle sensor data includes segmenting video data with a convolutional neural network to determine moving objects including vehicles.
4. The method of claim 1 , wherein the plurality of possible first vehicle path polynomials includes three or more path polynomials.
5. The method of claim 1 , wherein the planned path polynomial is determined by a deep neural network based on the identity of the first vehicle and the driving environment.
6. The method of claim 5 , wherein the driving environment is input to the deep neural network and output as a cognitive map that includes locations, speeds and directions for first vehicle.
7. The method of claim 5 , wherein the identity of the first vehicle is input to the deep neural network as data for hidden layers of the deep neural network.
8. The method of claim 1 , further comprising identifying the first vehicle based on vehicle sensor data including video sensor data.
9. The method of claim 1 , further comprising identifying the first vehicle based on vehicle-to-vehicle or vehicle-to-infrastructure communications.
10. A system, comprising a processor; and
a memory, the memory including instructions to be executed by the processor to:
determine a plurality of possible first vehicle path polynomials including predicted locations, speeds and directions of a first vehicle, and further determining respective probabilities of each of the path polynomials, based on (a) identifying the first vehicle according to at least one of a type of the first vehicle or a human driver of the first vehicle and (b) a driving environment in a second vehicle;
determine a second vehicle planned path polynomial for the second vehicle based on the determined probabilities corresponding to the predicted first vehicle path polynomial; and
operate the second vehicle based on the second vehicle a planned path polynomial.
11. The system of claim 10 , further comprising predicting the location, speed and direction for the first vehicle based on vehicle sensor data.
12. The system of claim 11 , wherein processing vehicle sensor data includes segmenting video data with a convolutional neural network to determine moving objects including vehicles.
13. The method of claim 1 , wherein the second vehicle planned path polynomial is based on the determined probabilities corresponding to the predicted first vehicle path polynomials being above a predetermined threshold.
14. The system of claim 10 , wherein the planned path polynomial is determined by a deep neural network based on the identity of the first vehicle and the driving environment.
15. The system of claim 14 , wherein the driving environment is input to the deep neural network and output as a cognitive map that includes locations, speeds and directions for the first vehicle.
16. The system of claim 14 , wherein the identity of the first vehicle is input to the deep neural network as programming data for hidden layers of the deep neural network.
17. The system of claim 10 , further comprising identifying the first vehicle based on vehicle sensor data including video sensor data.
18. The system of claim 10 , further comprising identifying the first vehicle based on vehicle-to-vehicle or vehicle-to-infrastructure communications.
19. A system, comprising:
means for controlling second vehicle steering, braking and powertrain;
computer means for:
determining a plurality of possible first vehicle path polynomials including predicted locations, speeds and directions of a first vehicle, and further determining respective probabilities of each of the path polynomials, based on (a) identifying the first vehicle according to at least one of a type of the first vehicle or a human driver of the first vehicle, and (b) a driving environment in a second vehicle;
determining a second vehicle planned path polynomial for the second vehicle based on the determined probabilities corresponding to the predicted first vehicle path polynomial and the means for controlling second vehicle steering, braking and powertrain; and
operating the second vehicle based on the second vehicle planned path polynomial for the second vehicle.
20. The system of claim 19 , further comprising predicting the location, speed and direction for the first vehicle based on vehicle sensor data.