Vehicle adaptive automatic driving decision-making method and system based on meta reinforcement learning

A technology of automatic driving and reinforcement learning, applied in control/regulation system, vehicle position/route/height control, non-electric variable control, etc., can solve problems such as unsafe decision-making, achieve flexibility, improve user experience, Safe and stable driving process

Active Publication Date: 2021-06-29
NANJING UNIV
View PDF8 Cites 8 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Taking automatic driving as an example, an automatic driving system that can run perfectly on a car, transplanted to a van may cause the system to make unsafe decisions due to changes in vehicle length, width and height

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Vehicle adaptive automatic driving decision-making method and system based on meta reinforcement learning
  • Vehicle adaptive automatic driving decision-making method and system based on meta reinforcement learning
  • Vehicle adaptive automatic driving decision-making method and system based on meta reinforcement learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0053] Below in conjunction with specific embodiments, the present invention will be further illustrated, and it should be understood that these embodiments are only used to illustrate the present invention and not to limit the scope of the present invention. The modifications all fall within the scope defined by the appended claims of this application.

[0054]In the present invention, the selection of driving behavior is based on the task information given by the task coding module and the state information given by the perception module. The task to be completed is to quickly and safely reach another point on the map under a certain vehicle condition. Obviously, if a certain driving scheme can drive reliably under the current road conditions, the scheme will be given a positive reward value; otherwise, it will be given a negative reward value. To maximize cumulative reward, we need to find the optimal mapping from environmental states and task encodings to driving behavior....

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a vehicle adaptive automatic driving decision-making method and system based on meta reinforcement learning, and the method comprises the steps: introducing a task coding module, recognizing current vehicle condition information from a vehicle driving track, coding the current vehicle condition information into vector representation, enabling the system to achieve the current vehicle condition, and timely adjusting the own driving strategy when the vehicle condition changes, thereby improving the driving efficiency, and the system is more robust and safer. In order to achieve better riding experience, reinforcement learning is used to solve the decision problem in the field of automatic driving. The system comprises a virtual environment module, a memory module, a sensing module, a coding module, a decision module and a control module. The reliability of the system is enhanced by adding task models which are extremely likely to be abundant to the virtual environment database; by changing the SAC reinforcement learning algorithm, the SAC reinforcement learning algorithm can make a decision based on the task coding module; and through a mode of maximizing mutual information between a task code and a sampling track, a task coding module can learn a task code containing rich information.

Description

technical field [0001] The invention relates to a vehicle self-adaptive automatic driving decision-making method and system based on meta-intensive learning, which is suitable for various types of vehicles with different specifications and does not need to re-learn a new automatic driving system for each vehicle. in the field of autonomous driving technology. Background technique [0002] Autonomous driving mainly needs to solve three core problems: state perception, path planning and the choice of driving behavior. At present, the problem of how to determine the state of the car, that is, "where am I", can be solved by using a variety of sensors; how to determine the overall path to the destination, that is, "how to get there", can use the current mainstream Dijkstra, A *, dynamic programming and other algorithms to solve. However, how to choose the most appropriate driving behavior according to the current state of the car, such as how to decide whether to pass at a cons...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): B60W60/00G05D1/02
CPCB60W60/0011G05D1/0221
Inventor 章宗长俞扬周志华胡亚飞徐峰
Owner NANJING UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products