Intersection motion planning method and device of automatic driving equipment, and electronic equipment

Abstract

The invention discloses an intersection motion planning method of automatic driving equipment, belongs to the technical field of automatic control, and helps to save computing resource consumption during motion planning. The intersection motion planning method disclosed by the embodiment of the invention comprises the following steps of: acquiring a real-time joint state including state data of automatic driving equipment and state data of obstacles around the automatic driving equipment; determining a predefined event of joint state matching, wherein the predefined event includes: a controllable event indicating that an action needs to be taken; in response to the fact that a predefined event matched with the joint state is the controllable event, carrying out mapping processing from an event space to an action space on the joint state through a pre-trained reinforcement learning network, thus obtaining an action vector matched with the joint state; and performing motion planning on the automatic driving equipment based on the obtained action vector, thus saving calculation resource consumption during intersection motion planning, and improving the intersection motion planning efficiency of the automatic driving equipment.

Description

technical field [0001] The embodiments of the present application relate to the technical field of automatic control, and in particular to a method, device, electronic device, and computer-readable storage medium for intersection motion planning of automatic driving equipment. Background technique [0002] The motion planning module is the core technology module in the automatic driving system. The motion planning module controls the motion direction and speed of the automatic driving equipment based on the information obtained by the perception and positioning module, and plans an efficient motion path that meets the safety, comfort, and vehicle dynamic constraints. One way of implementing the motion planning module in the prior art is to use deep learning methods to obtain high-level semantic information of collected environmental images, and use reinforcement learning methods to complete end-to-end real-time scene path planning from the environment. In the motion plannin...

AI Technical Summary

Problems solved by technology

In the motion planning scheme in the prior art, the learning process is slow during the model training process, and the model is not easy to converge

Moreover, when processing the collected environmental images in the motion planning process, a large number of features need to be extracted from the images. For the scene where the automatic driving equipment decides the road motion state, it contains many redundant states, which wastes the calculation of the automatic driving equipment. resource

Images