COOPERATIVE ATTACK STRATEGY OF UNMANNED AERIAL VEHICLES IN ADVERSARIAL ENVIRONMENT

摘要

This paper addresses the problem of risk in adversarial environments and presents a new attack model in which the cooperative action is considered. The actions of each UAV not only get benefit from attacking targets but also reduce the risk in the environment for other UAVs. The implications of the survival probabilities and the time discount factor in this model are analyzed and the results will help to build the optimal attack strategy. An improved genetic algorithm is applied to quickly build the optimal strategy. Simulation results show that this algorithm provides better solutions than genetic algorithm.