A approach guidance method for carrier-based aircraft based on deep reinforcement learning

Abstract

The invention discloses a shipboard aircraft approach guiding method based on deep reinforcement learning, and the method comprises the following steps: firstly carrying out the modeling of a shipboard aircraft and an aircraft carrier, and constructing an intelligent agent training environment; secondly, constructing a deep reinforcement learning guidance intelligent agent, and designing a state space and a decision action space of the intelligent agent; then, setting a reward function according to a shipboard aircraft approach success condition; then, setting initial attitudes of a shipboardaircraft and an aircraft carrier in the guidance scene, and training an intelligent agent by adopting a deep reinforcement learning method; and finally, accurately guiding the shipboard aircraft to reach a final approach point by using the trained intelligent agent. The method can be applied to intelligent guidance of the shipboard aircraft; the intelligent agent with the command and control capability is used for assisting a commander in commanding, the shipboard aircraft is guided to reach the final approach point from any posture, the problems that in the approach process, the approach efficiency is low and consumed time is long due to the fact that the shipboard aircraft waits for approach in a fixed air route are solved, and the approach process is more efficient and reliable.

Description

technical field [0001] The invention belongs to the technical field of computer application and artificial intelligence, and especially designs a carrier-based aircraft approach guidance method based on deep reinforcement learning. Background technique [0002] The carrier-based aircraft is an important guarantee for the combat effectiveness of the aircraft carrier, and whether the carrier-based aircraft can reach the approach point efficiently and reliably is one of the most important technical conditions to ensure the combat effectiveness of the aircraft carrier. The existing approach guidance technology requires carrier-based aircraft to wait on a fixed route and approach sequentially, resulting in low efficiency and time-consuming approach guidance, which cannot meet the requirements of efficient and reliable approach process. The present invention proposes a carrier-based aircraft approach guidance method based on deep reinforcement learning. This method can use the dee...

AI Technical Summary

Problems solved by technology

[0003] The purpose of the present invention is to solve the problem of low efficiency and time-consuming approach guidance technology of the existing carrier-based aircraft, and propose a method of approach guidance for carrier-based aircraft based on deep reinforcement learning, which can enable the carrier-based aircraft to be able to Efficiently and reliably reach the approach point, improving the approach guidance efficiency of carrier-based aircraft

Images