Automatic driving decision-making method and system

Abstract

The embodiment of the invention provides an automatic driving decision-making method and belongs to the automatic driving field. The automatic driving decision-making method includes the following steps that: the input data and output data of a decision-making model of a vehicle driving system are acquired; the decision-making model is trained according to the input data and output data of the decision-making model; and the trained decision-making model is made to output decision-making behaviors about driving behaviors according to the input data of the decision-making model. The automatic driving decision-making method of the present invention can adapt to a complicated and varying driving environment, thereby realizing automatic driving that can be directly applied to an unmanned vehicle.

Description

technical field [0001] The present invention relates to the field of automatic driving, in particular to an automatic driving decision-making method and system. Background technique [0002] In recent years, unmanned driving-related technologies have received widespread attention. Among them, how to make decisions on the driving behavior of unmanned vehicles based on surrounding obstacle information and road information is very important during the driving process of unmanned vehicles. [0003] The existing decision-making algorithms for unmanned vehicles usually use a decision-making method based on finite state machines, that is, the decision-making rules for unmanned vehicles are obtained based on vehicle driving experience, and the expected driving behavior is obtained based on road information and surrounding obstacle information as state machine input. . [0004] However, the inventors of the present application found in the process of implementing the present inventi...

AI Technical Summary

Problems solved by technology

[0004] However, the inventors of the present application found in the process of implementing the present invention that the decision-making method based on the state machine needs to give the decision-making rules in advance, which cannot deal with the lack of decision-making rules, and it is usually very complicated to extend the decision-making rules. This method cannot adapt to complex and changeable traffic scenarios

At the same time, the decision-making method based on machine learning is applied to the assisted driving system, but the assisted driving system relies more on the driver's behavior to give the final decision result, and cannot be directly applied to the decision-making system of driverless cars.

Images