In response to the long optimization time and high energy consumption faced by traditional particle swarm optimization algorithm in global path planning for autonomous underwater vehicle,this paper proposes an improved T-dis-tribution fireworks-particle swarm optimization algorithm(TFWA-PSO),this algorithm integrates the efficient global search capability of the fireworks algorithm with the rapid local optimization characteristics of the particle swarm optimization algo-rithm.In the mutation stage,an adaptive T-distribution mutation is proposed to expand the search range,and it is theoretical-ly demonstrated that this explosive mutation approach enables individuals to enhance their search ability near the local opti-mal solution.In the selection stage,a fitness selection strategy is proposed to eliminate individuals with poor fitness,solving the problem of the traditional fireworks algorithm's tendency to lose excellent individuals,and comparing the convergence speed between the improved T-distribution fireworks algorithm and the traditional fireworks algorithm.The improved algo-rithm's explosion,mutation operations,and selection strategy are integrated into the particle swarm algorithm.The velocity update formula of the particle swarm algorithm is improved,while the convergence proof of the improved algorithm is proved theoretically.The simulation results indicate that the TFWA-PSO can effectively plan the shortest path.Compared to the given intelligent optimization algorithms,TFWA-PSO on average reduces the time to find the optimal path by 24.72%,lowers energy consumption by 17.33%,and decreases the average path length by 16.96%.
维普
万方数据