Method for unmanned aerial vehicle autonomous counter-reconnaissance and enemy attack avoidance

Abstract

The invention discloses an unmanned aerial vehicle autonomous counter-reconnaissance and enemy attack avoidance method, which comprises the steps of constructing a simulator based on Unity 3D, and constructing a battle scene of attacks of both enemy and my parties; performing limited round confrontation with a small number of opponents written according to rules through a reserved human player interface of the simulator, and collecting sample data of a solution for the unmanned aerial vehicle to avoid enemy counter-reconnaissance and attack; learning by utilizing the collected escape data of the human player, and learning through imitation learning in reinforcement learning to obtain a pre-training model; based on the pre-training model, realizing escape ability learning in the simulator through a reinforcement learning algorithm PPO; and testing the trained unmanned aerial vehicle avoidance reconnaissance and attack model and a human player, and migrating the model to a real environment. The unmanned aerial vehicle autonomous attack avoidance solution provided by the invention is of great application significance in a real scene, can train a good unmanned aerial vehicle attack avoidance strategy on the premise of low cost, and can be practically applied to the field of unmanned aerial vehicle flight control.

Description

technical field [0001] The invention relates to a method for autonomous anti-reconnaissance and avoiding enemy attacks of unmanned aerial vehicles, which mainly solves how to effectively avoid enemy anti-reconnaissance and attacks during the flight process of reconnaissance unmanned aerial vehicles, and belongs to the technical field of unmanned aerial vehicle flight control. Background technique [0002] With the advancement of science and technology and the improvement of the technical level of drones, drones are becoming more and more important in the military, and are widely used in military reconnaissance, combat and other fields. In the process of carrying out military reconnaissance missions, UAVs often face the possibility of being detected or even attacked by opponents; at the same time, the complexity and changes of the environment itself also bring extreme dangers for UAVs to avoid reconnaissance or attacks. Big challenges, such as weeds, shrubs, and buildings wit...

AI Technical Summary

Problems solved by technology

[0006] Purpose of the invention: In view of the problems in the prior art that the control of UAVs is relatively difficult and the generalization ability of the rule-based escape scheme is poor, the present invention provides a method for autonomous anti-reconnaissance and evasion of enemy attacks by UAVs. It is an effective and practical approach to build a simulator, and first learn a better policy through a reinforcement learning algorithm in the simulator, and then migrate it to the real environment

Images