A real-time collision avoidance method for redundant dual-arm robots in an open operational environment

Abstract

Due to the structural resemblance of redundant dual-arm robots to human arms, they are widely employed to replace humans in open operational environments. Addressing safety concerns related to the autonomous operations of redundant dual-arm robots in open environments, this paper proposes a real-time collision avoidance method. Firstly, an avoidance direction adjustment algorithm is designed based on the avoidance function method, providing a collision avoidance formulation for the robot control point. Secondly, an obstacle classification algorithm is devised to categorize obstacles into robot body obstacles and end-effector obstacles, and the collision avoidance strategy of redundant dual-arm robots is designed. Subsequently, a collision avoidance penalty factor is introduced based on the proximity between the end-effector and the target point, ensuring the convergence of the joint velocity. Finally, a novel collision avoidance formulation for redundant manipulators is presented, further extended under dual-arm coordinated tasks. Numerical simulations and physical experiments demonstrate that the proposed method can achieve self-collision avoidance for redundant dual-arm robots and dynamic/static obstacle avoidance in dual-arm coordinated tasks, with smooth collision avoidance maneuvers. The research results provide safety guidelines for autonomous operations of redundant dual-arm robots in open operational environments.