Performance evaluation of particle swarm intelligence based optimization techniques in a novel AUV path planner

Over years of development, many optimization techniques have been proposed for the path planning of the Autonomous Underwater Vehicle (AUV). The development in swarm intelligence optimization, particularly the particle swarm optimization (PSO), has significantly improved the performance of the AUV path planner. This study presents 12 variants of particle swarm intelligence (PSI)-based algorithms, which were applied to evaluate their performances in solving the optimal path planning problem of an AUV operating in 2D and 3D ocean environments with obstacles and non-uniform currents. Throughout the structure of the optimization problem, the practicability of the path planning algorithms were considered by taking into account the physical limitations of the AUV actuations. To compare the performances of these PSI-based algorithms, extensive Monte Carlo simulations were conducted to evaluate these algorithms based on their respective solution qualities, stabilities and computational efficiencies. Ultimately, the strengths and weaknesses of these algorithms were comprehensively analyzed, in order to identify the most appropriate optimization algorithm for AUV path planning in dynamic environments.