Abstract: Under the research background of collaborative multi-aircraft and multi-target air combats, combined with the actual combat constraint conditions and the threat assessment functions on both sides, a collaborative air combat target decision simulation model is established for complex and changeable battlefield situations, which can reflect the priority of fire attack. To solve the decision scheme quickly and accurately, an improved multi-agent particle swarm optimization algorithm is proposed by introducing the interaction mechanism of the multi-agent theory into particle swarm optimization algorithm; and the neighborhood cooperation operator, mutation operator and self-learning operator for the agent are designed respectively. The simulation results show that the method can work out a reasonable and effective decision scheme, and has a good real-time simulating performance.

Key words: collaborative air combat, multi-target decision, multi-agent system, particle swarm optimization