An ant colony path planning optimization based on opposition-based learning for AUV in irregular regions

Aiming at the problems of incomplete path coverage and path redundancy in Autonomous Underwater Vehicle (AUV) path planning, an Ant Colony Path Planning Optimization Based on Opposition-Based Learning (ACPPO-OBL) is proposed. Firstly, Opposition-Based Learning (OBL) is introduced during the initialization phase of the ant colony. Moreover, the theoretical proof that ant colonies can be distributed near the optimal ant colony has also been proposed, indicating that the ACPPO-OBL algorithm has enhanced global search ability. Secondly, the coefficient for pheromone evaporation is revised. Besides, the proposed method involves a global pheromone update incorporating both best and worst reward mechanisms. Furthermore, it has been theoretically proven that the ACPPO-OBL algorithm has upper and lower bounds on the total pheromone concentration when searching for the optimal path. Additionally, an adaptive coefficient is incorporated into the heuristic function. The theoretical proof of the convergence of ACPPO-OBL has been established. As demonstrated in simulation experiments, ACPPO-OBL increases path coverage rates by 2–6 \(\%\) $\%$ and reduces path lengths by 6–11 \(\%\) $\%$ compared to ECDM planning. The ACPPO-OBL can be applied to cover irregular areas of various shapes and provides better coverage, improving the efficiency and stability of full-coverage paths in irregular areas.