Path Planning of Robot Fish Based on RRT and Artificial Potential Field Algorithm

In the underwater robot fish water polo 2v2 global visual group competition, it is crucial for the robot fish to select a feasible path from the initial position to the target point in order to better accomplish the ball topping task. To address the randomness, blindness, and inefficiency inherent in the traditional Rapidly-exploring Random Tree (RRT) algorithm, the concept of artificial potential fields was introduced. This approach guides the random tree to grow towards the target direction and avoid dynamic obstacles, By doing so, effectively reducing the randomness in path exploration for the robot fish and improving planning speed. Furthermore, by integrating Bezier curves to optimize the planned path, the smoothness of the path was enhanced, effectively preventing rapid directional changes during robot fish movement and reducing energy consumption. Through simulation experiments and practical tests, it has been demonstrated that the improved algorithm can accomplish tasks more accurately and efficiently.