Vibration Damping Control of Cable-Climbing Robots Based on PPO Algorithm

To address the climbing instability of cable-climbing robots in environments with cable vibrations, a vibration reduction control method based on the PPO (Proximal Policy Optimization) algorithm is proposed. Firstly, based on a nitrogen spring-magnetorheological (MR) damper with variable stiffness and damping coupled loading mechanism, a BP neural network output force model is established. Simulation results show that the error of the designed output force model is within 6.6N, demonstrating the accuracy of the model. Then, the workspace and state space for the robot output under the PPO algorithm are determined, and the reward function and network structure are designed to ensure effective exploration during the learning process. Finally, a simulation environment is built in MATLAB Simulink. Simulation results indicate that, with a cable vibration frequency of 10Hz and amplitude of 4mm, the PPO algorithm improves the climbing speed stability of the robot by 31.6% compared to PID control.