Intelligent Robotic Arm Path Planning (IRAP2) Framework to Improve Work Safety in Human-Robot Collaboration (HRC) Workspace Using Deep Deterministic Policy Gradient (DDPG) Algorithm

Abstract Industrial robots are widely used in manufacturing systems. The places that humans share with called human-robot collaboration (HRC) workspaces. To ensure the safety HRC workspaces, a collision-avoidance system is required. In this paper, we regard as problem during robot action trajectory design and propose an intelligent robotic arm path planning (IRAP 2 ) framework. IRAP framework based on deep deterministic policy gradient (DDPG) algorithm because typical continuous control dynamic environment, DDPG well suited for such problems. test framework, have studied workspace which size larger than humans. At first, applied physics engine to build virtual including digital models of human. Using environment model, further trained agent model using algorithm. can optimize motion avoid collision