Control method and device based on reinforcement learning, and controlled motion equipment

Abstract

The invention provides a control method and device based on reinforcement learning, and controlled motion equipment. The method comprises the following steps: obtaining the situation information of the controlled motion equipment, and judging whether a deviation angle is greater than an angle threshold value or not; and if it is judged that the deviation angle is larger than an angle threshold value, inputting the situation information into an angle capture network obtained through pre-training, outputting a first action of controlled motion equipment through the angle capture network so as tocontrol the controlled motion equipment to enter a conical space, and controlling the controlled motion equipment to approach the target through the distance capture network in the conical space. Theinteraction space of the distance capture network is the conical space, and then only the coordinates in the conical space are used as the training samples to train the distance capture network, so that the training difficulty is reduced, and the fighter aircraft is quickly and accurately controlled to autonomously maneuver to the attack area.

Description

technical field [0001] The invention belongs to the technical field of automatic control, and in particular relates to a control method, device and controlled motion equipment based on reinforcement learning. Background technique [0002] In the DRL (Deep Reinforcement Learning, Deep Reinforcement Learning)-based control method of the controlled object equipment (such as a fighter plane), the autonomous maneuvering of the fighter plane to the target is a prerequisite for the fighter plane to effectively strike the target. [0003] However, the interaction space of fighter jets is large, and it is difficult to accurately control the fighter jets to autonomously maneuver to the target area when controlling the fighter jets to maneuver to the target based on DRL in the prior art. Contents of the invention [0004] In view of this, the object of the present invention is to provide a control method, device and controlled motion equipment based on reinforcement learning, which i...

AI Technical Summary

Problems solved by technology

[0004] In view of this, the purpose of the present invention is to provide a control method, device and controlled motion equipment based on reinforcement learning, which is used to solve the problem of accurately controlling the fighter to maneuver autonomously to the target area in the existing scheme based on DRL. Problems in the target area

Images