Collision avoidance strategy based on virtual body deformation for path planning of serial industrial robot

Collision-free path planning is a crucial capability for industrial robot in its application. In this paper, a collision avoidance strategy based on virtual body deformation (VBDCAS) is proposed to improve the planning efficiency. The main idea of VBD-CAS is to realize the obstacle avoidance according to the allowable deformation space of the robotic arm relative to the obstacle. The obstacle avoidance evaluation model with spherical obstacle is established through the limit motion analysis of robotic arms. Then, a particle swarm optimization with VBDCAS (PSO-VBD-CAS) algorithm is developed for the collision-free path planning by integrating VBD-CAS into PSO algorithm. PSO-VBD-CAS algorithm uses obstacle avoidance evaluation model to judge weather the collision is happened between robotic arm and obstacles. When the joint angles of robot is located in the allowable deformation space of robotic arm, the robotic arm does not collide with obstacles. The proposed algorithm can directly perform the collision judgment without repeatedly performing forward kinematics solution of robot and shortest distance calculation from robotic arm to obstacles. Two simulations and an experiment are used to verify the validity of the proposed method. The validation results demostrate that the proposed VBD-CAS can effectively realize the collision detection, meanwhile the PSO-VBD-CAS algorithm has a better efficiency compared to distance-based obstacle avoidance path planning methods.