Intelligent Assignment Strategy for Multi-Target Adversarial Interception

Abstract

The multi-missile confront multi-target is a classic target allocation issue in the combat scenario of multiple missiles intercepting multiple maneuvering targets. Traditional algorithms lack environmental assessment model, train quality, and indicator function in the adversarial environment. To this end, this paper aims to propose an intelligent assignment strategy which contains indicator function and evaluation model. Then, an indicator function and an evaluation model considering the miss distance, threat situation, and the number of specified interception targets are introduced into the reinforcement learning algorithm. The local and global reward functions are introduced to improve the training convergence and efficiency in the multi-missile multi-target confrontation scenario. Finally, simulation results are designed to check on advantage of intelligent allocation strategy.