Path Planning for Multi-Arm Manipulators Using Soft Actor-Critic Algorithm with Position Prediction of Moving Obstacles via LSTM

This paper presents a deep reinforcement learning-based path planning algorithm for the multi-arm robot manipulator when there are both fixed and moving obstacles in workspace. Considering problem properties such as high dimensionality continuous action, proposed employs SAC (soft actor-critic). Moreover, order to predict explicitly future position of obstacle, LSTM (long short-term memory) is used. The SAC-based developed using LSTM. In show performance algorithm, simulation results GAZEBO experimental real manipulators presented. experiment that success ratio generation arbitrary starting goal points converges 100%. It also confirmed successfully predicts obstacle.