Obstacle Avoidance Path Planning for Spherical Joint Actuator Based on Improved APF-Bi-RRT* Algorithm

Abstract

To avoid collisions between an end tool driven by a novel spherical joint actuator and surrounding obstacles during operations in static scenarios such as assembly and spraying, which may lead to system malfunctions, this paper proposes a method that integrates an improved artificial potential field (APF) and a dynamic target biasing strategy with the bidirectional rapidly-exploring random tree star (Bi-RRT*) algorithm. This approach, termed the Improved APF-Bi-RRT* algorithm, is combined with a novel collision detection method to achieve efficient path planning. Firstly, collision detection is reframed from analyzing contact between the end tool and obstacles to directly defining the unreachable region of the end point. On this basis, an environment map is reconstructed by leveraging the spherical motion characteristics of the end point, thereby satisfying workspace constraints while reducing the complexity of subsequent obstacle avoidance algorithms. Then, the Bi-RRT* algorithm is applied for path search on this map, and the improved APF and dynamic target biasing strategy are integrated into the search process to minimize redundant nodes and accelerate convergence. Finally, simulation and experimental results demonstrate that the proposed method not only ensures reliable obstacle avoidance for the end tool driven by the spherical joint actuator in static known obstacle environments, but also reduces planning time and path length compared with the RRT*, P-RRT*, Bi-RRT*, and APF-Bi-RRT* algorithms.