Optimal Path Planning for Underwater Robots Based on Improved Ant Colony Algorithm

Abstract

When applying traditional ant colony algorithm in the path optimization of underwater robots, there are several problems such as slow convergence speed and poor optimization effect. In this paper, an improved ant colony algorithm was proposed. The improved algorithm enhances the pheromone concentration in the core area of the grid map in the first inquiry stage, which can improve the confluence efficiency of the algorithm. In order to reduce the number of turns of the ants and make the path smoother, the corner heuristic function was added to the state transition probability equation. The position of the target point was added to the heuristic function to boost the target point's guiding influence on the ant colony. In the pheromone update part, the allocation strategy of the wolf pack algorithm was introduced to strengthen the pheromone of the optimal path and at the same time limit the pheromone concentration to reduce the generation of local optimal solutions. In the MATLAB simulation verification, the improved ant colony algorithm plans a shorter path length and fewer turns. The algorithm can effectively avoid obstacles, has better global optimization, and avoids the energy loss of underwater robots. This paper verifies the feasibility and superiority of the improved ant colony algorithm in static path planning.