Dynamic Obstacle Avoidance Algorithm for Robot Arm Based on Deep Reinforcement Learning

Abstract

A dynamic obstacle avoidance planning algorithm based on deep reinforcement learning is proposed for rigid manipulators. After the neural network interacts with the environment and learns, it can give real-time action strategies to guide the manipulator to avoid dynamic obstacles. This paper proposes a new state space description method suitable for manipulators and dynamic environments, and designs the corresponding collision detection method and reward value calculation function for this state description method. The test results in the simulation environment demonstrate the effectiveness of the method.