Precise robust control method for charging manipulator considering dynamic disturbances

Charging manipulator control technology is an important means of improving the efficiency and convenience of electric vehicle charging and reducing charging risks. However, it faces challenges such as difficulty in adapting to dynamic disturbance environments, slow convergence speed, and susceptibility to becoming stuck in local optima. To address these challenges, this study proposes a precise robust control method for charging manipulators considering dynamic disturbances. First, a manipulator model is constructed, and the dynamic disturbance value of the manipulator is estimated based on fuzzy logic. Second, a single-joint Markov Decision Process (MDP) model was constructed, and a path dynamic planning method for the manipulator based on the back-propagation Deep Deterministic Policy Gradient (DDPG) was proposed. Through back-propagation training, disturbance-aware experience replay, and local-global coordinated parameter updating and convergence judgment, global optimization, and precise robust control of the charging manipulator are achieved. Finally, the performance of the back-propagation DDPG algorithm was verified by simulation. The simulation results show that the proposed method increases the convergence speed of the manipulator by 61.54% and the success rate of task completion by 18.94%, effectively enhancing the precision and robustness of charging manipulator control.