An Improved Multi-objective Path Planning Algorithm and its Application to Tennis Ball Collection Robots

The tennis collection robot can autonomously collect scattered tennis balls on the court, effectively reducing the workload of athletes and improving training efficiency. It has great application value in the field of sports. In this paper, we propose a multi-objective path planning scheme based on the improved A * and ACO (Ant Colony) algorithms for the ball-picking path planning problem of the tennis collecting robot. By adding a dynamic weight to the heuristic function of A *, the running rate of the algorithm is improved, and the multi-objective solution space with obstacle avoidance is obtained. On the basis of this solution space, the sequence of multi-objective traversals with fixed start and end points is solved using the improved ACO algorithm. The multi-objective ball pickup path is determined. In addition, this paper realizes the recognition and localization of tennis targets based on binocular camera and Yolov5s target detection algorithm. And using PID fixed-point navigation and visual servo strategy, the tennis ball pickup task was accomplished. Through experimental verification, the scheme proposed in this paper is robust and scalable. The tennis ball collection robot can work autonomously and collect tennis balls continuously and efficiently.