Path Planning of an Automatic Piecing Robotic Arm for Ring Spinning Based on an Adaptive RRT-Connect Algorithm

Ring spinning machine is widely used in the textile industry, yet yarn breakage remains a persistent problem that degrades production efficiency and product quality. To address this issue, a robotic arm and its path-planning algorithm are employed to perform the yarn piecing task. This paper presents an adaptive RRT-Connect path-planning algorithm for yarn piecing, developed for a robotic arm equipped with a custom-designed end connector. A bounding-box method is employed to model collisions between the arm and surrounding obstacles, thereby reducing the complexity of collision detection. An adaptive sampling strategy replaces conventional random sampling to reduce invalid samples and accelerate forward search. During tree expansion, a dynamic bidirectional tree-expansion strategy is introduced to speed up the connection process between the two trees and to limit exploration of irrelevant regions. Simulations demonstrate that the proposed adaptive RRT-Connect method outperforms conventional path-planning algorithms in the yarn piecing task and substantially improves the piecing success rate.