An improved particle swarm optimization based on age factor for multi-AUV cooperative planning

Multi-AUV planning of cooperative path represents the intelligence level of AUV formation system, which has attracted significant attention due to its flexibility and robustness. To enhance the speed and exploration capability of the multi-AUV autonomous path planner, this paper proposes an Age-based Multi-population Particle Swarm Optimization (AMPSO) for three-dimensional path planning of multiple AUVs. Firstly, reverse chaotic mapping is employed to enhance the initial distribution, enabling them to better adapt to the optimization process. In addition, the introduction of multiple populations helps to enhance the diversity within the population, thereby accelerating the convergence rate. Taking into account the search ability, the concept of an “age” coefficient is proposed to optimize the velocity updating formula to reduce the effect of inferior particles. Furthermore, a velocity-independent updating mode is suggested to enhance the population diversity by selecting different update modes. To evaluate the performance of the proposed algorithm, Monte Carlo simulations are conducted for AUV formations across two tracking tasks within the underwater environment established in this paper. The results demonstrate that the proposed algorithm outperforms other commonly used algorithms in terms of both speed and optimality, so the proposed AMPSO can be used for multi-AUV cooperative planning.