A master-slave teleoperation system control model based on model predictive control with fast dynamic response

Abstract

Robot technology is advancing with constantly expanding applications. Teleoperation system demonstrates its potential to extend human capabilities in unstructured and hazardous environments as an important branch of robotics. However, there are still issues involving insufficient position tracking accuracy and dynamic response capabilities. This article proposes a master-slave teleoperation system control model based on model predictive control (MPC) with a real-time dynamic adaptive parameter adjustment algorithm. We establish a model based on the integral barrier Lyapunov function and construct a dynamic optimization model based on MPC. Then we carry out rolling optimization and feedback correction at each moment to feed the predicted values generated by MPC back to correct the actual tracking trajectories in real-time. Additionally, we utilize an adaptive neural network and robust term compensation to adjust dynamic parameters and eliminate the effects of time delays, external disturbances, and model uncertainties. Theoretical simulation experiments are conducted on the Simulink platform, analyzing position tracking error and dynamic response capabilities of the master-slave robot respectively to verify the proposed model. The results indicate a reduction in the position tracking errors of the master with 10% and slave with 11%, while also showing an improvement in their dynamic response capabilities. The speed error ranges of the master and slave are shortened with 6% and 7% respectively.