Chaos-Based Inertia Weight Particle Swarm Optimization for Path Planning of Unmanned Surface Vehicles

Abstract

Unmanned surface vehicles (USVs) play a vital role in marine construction, such as maritime patrol and search-rescue operations. However, USVs face complex challenges when performing tasks in path planning. An adaptive chaotic-enhanced particle swarm optimization (ACA-PSO) algorithm is presented in this paper to increase the efficiency of static path planning for USVs in intricate maritime environments. The proposed ACA-PSO algorithm addresses the dual challenges of path optimality and navigational safety while considering path length, smoothness, safety, and convergence speed. First, a sine chaotic sequence for initialization is introduced, which enhances global search capabilities and effectively escapes local optima. Second, an adaptive inertia weight decay strategy is proposed to adjust the weight dynamically to enhance convergence efficiency and solution quality. Third, sinusoidal acceleration learning factors are considered to alleviate the limitations of static cognitive and social learning factors. Fourth, a historical velocity weighting strategy is developed to improve trajectory smoothness and convergence robustness. Finally, extensive simulation results confirm that the proposed ACA-PSO algorithm significantly outperforms conventional PSO and other optimization algorithms in path planning effectiveness, convergence efficiency, and operational stability. Specifically, in optimal path distance, the proposed ACA-PSO algorithm achieves 2.4% and 4.9% improvements compared with traditional PSO and simulated annealing (SA) algorithms, respectively. The proposed ACA-PSO algorithm also demonstrates a 29.8% better grayscale value in path smoothness compared with the SA algorithm. Moreover, with a much lower standard deviation of 0.201 × 10^-2, the ACA-PSO algorithm is superior in stability.

Keywords

Unmanned surface vehicles (USVs) / Particle swarm optimization (PSO) / Path planning / Chaos sequences / Adaptive inertia weights

RIGHTS & PERMISSIONS

Harbin Engineering University and Springer-Verlag GmbH Germany, part of Springer Nature