Path Planning for Robotic Fish Based on Improved PRM Algorithm Fused with DWA Algorithm

With the dramatic development of remote operated vehicle (ROV), how to ensure robotic fish safety operation has been a growing concern. This study addresses the challenge posed by the traditional PRM algorithm’s inability to adequately optimize paths due to a lack of sufficient sampling points in robotic fish path planning. To mitigate this limitation, we propose an enhanced PRM algorithm that integrates seamlessly with the Dynamic Window Approach for robotic fish path planning. By introducing the artificial potential field method to apply forces to sampling points on obstacles and optimizing them, and combining the DWA algorithm of local path planning to select paths according to the evaluation function, more effective dynamic obstacle avoidance can be achieved. The experimental results show that the algorithm improves the utilization of sampling points, reduces path distance by 16.48%, and saves time by 26.76%. Prove that the path planned by this algorithm is shorter and smoother, improving the efficiency of the robot’s work.