Global path planning for unmanned surface vehicles based on improved ant colony optimization and waypoint refinement

To overcome the deficiencies of traditional ant colony optimization (ACO) in pheromone updating, local optima convergence, and path planning safety, this study proposes a global path planning algorithm based on improved ACO and turning-point refinement. The heuristic function is improved using the reciprocal of the Euclidean distance between current path nodes and the destination, along with balancing parameters for iteration number, search quality, and efficiency, thereby enhancing global and local search capabilities while avoiding local optima. An adaptive pheromone evaporation coefficient is designed by utilizing characteristics of cosine function to dynamically adjust the convergence of the proposed ant colony optimization in its early and late stages. Considering the complexity of maritime environments and practical navigation requirements, a grid-based navigation environment is constructed. An obstacle-adjacent node detection method and fixed-point approximation algorithm is proposed for turning point refinement to improve navigation safety and ensure optimized paths better conform to maritime practice. Simulation experiments demonstrate that, compared with traditional ACO and other improved algorithms, the proposed algorithm shortens the average path length by approximately 39% and reduces the average iteration number by 79%, significantly improving solution quality and convergence efficiency. It effectively alleviates issues of insufficient search directionality and susceptibility to local optima. These results verify the reliability of the proposed approach for global path planning of unmanned surface vehicles and its high efficiency in redundant waypoint optimization, thereby providing effective decision support in practical applications.