Research on USV Route Planning Based on Simulated Annealing-Chaos Adaptive Particle Swarm Optimization Algorithm

In this paper, a simulated annealing adaptive particle swarm optimization algorithm based on the Logistic chaotic mapping model (SA-CAPSO) is proposed to solve the problems of premature, long initialization time, insufficient anti-jamming adaptability, and low convergence speed and accuracy of traditional particle swarm optimization (PSO) for unmanned surface vehicle (USV) route planning in a relatively complex environment. First, in this method, the adaptive coefficient is used to replace the original intrinsic coefficient to improve the velocity distribution of particles. Secondly, the chaos map strategy is introduced to improve the initial division of particles by using its sensitivity to the initial value. Finally, the simulated annealing algorithm (SA) is integrated to improve the optimality of the solution by taking advantage of its probabilistic global optimization characteristics. To evaluate the SA-CAPSO algorithm, we compared the route planning results with the standard Particle Swarm Optimization (PSO), Adaptive Particle Swarm Optimization (APSO), and Chaos Adaptive Particle Swarm Optimization (CAPSO). The experimental results show that the SA-CAPSO algorithm can obtain the planned track curve with good convergence accuracy and speed.