第 5 卷第 1 期 无人系统技术 202rithm based on cooperative particle swarm Optimization (CPSO) was proposed to track underwater targets by sea-air cross-domain unmanned vehicle cluster in complex water environment. In the CPSO algorithm, firstly, the flight path is planned for the UAVs, and obstacles in the water navigation area are detected during the UAV flight. Then, a complete surface obstacle distribution map is generated for unmanned surface vehicles (USVs) cluster. Subsequently, CPSO algorithm was used to plan the cooperative navigation path for USVs, and the underwater navigation environment map was generated for autonomous underwater vehicles (AUVs) during the cluster navigation. Finally, CPSO algorithm plans the underwater navigation path for AUVs, and AUV completes the underwater target tracking task. The simulation results show that the CPSO algorithm plans a valid path for the sea-air cross-domain unmanned vehicle cluster, and realizes the function of unmanned vehicle cluster cooperative operation