An Adaptive Control Method for Robot Constant Force Polishing Device

Abstract

Grinding and polishing of product parts are difficult points for industrial robots to be used in actual manufacturing. In constant force grinding process, the quality of the grinding process depends largely on whether the normal contact force can be kept constant during the grinding process. For this reason, an adaptive control method combined with fuzzy logic for the robot grinding and polishing device is proposed in this paper. In this method, an improvement of the single neuron proportional integral derivative (PID) control method is adopted and applied to the grinding and polishing device of industrial robots. Through simulation, it is found that this method can effectively reduce the adjustment time in contact force control. Aiming at the situation that the device is interfered many times, a non-linear function of the variable learning rate benchmark is proposed to improve the robustness of the system.