Application and performance analysis of deep reinforcement learning algorithm in path planning of dual-arm robots

In this paper, path planning of dual-arm robot is the research topic, and the application and performance analysis of deep reinforcement learning algorithm are discussed. Firstly, the basic principles of path planning for dual-arm robots are briefly introduced, including kinematics and dynamics models and common path planning algorithms. Then, the application of deep reinforcement learning algorithm in environment modeling, state space design, reward function tuning and training optimization is analyzed. Finally, the performance of deep reinforcement learning path planning is compared and analyzed in detail through path quality, computational efficiency, robustness and generalization ability.