A Method on Impedance Control of Robots Based on the Neural Network

In this paper a method to implement compliance control of robots is presented. Under the condition of unknowing the robotpsilas precise model, the robot is approximately decoupled into a number of independent SISO linear subsystems. An ANN is designed to construct a inverse system and the well-trained ANN inversion is cascaded with the manipulator for decoupling. For the above decoupled position system, a control algorithm based on the target impedance is presented to regulate the impedance of the robot and perform compliance control. Simulation and experimental results show good performance of decoupling and real-time tracking any arbitrary trajectories and validity of this method for compliance control of robots.