Optimal Predictive Delay Compensation for Robot Networked Control System

In order to solve the problem of random delay in robot networked control system, an optimal predictive compensation control method is designed. The state observer is used to observe the joint position information of the robot at the current time, and it is sent to the control predictive generator. The predictive generator forecasts and calculates the control input at the later time, and optimizes the predictive result of the control input by rolling optimization. Finally, simulations are conducted on a two-joint robot networked control system. The simulation results show that the method can effectively compensate for the influence of random delay and packet disorder on the position tracking performance of the system.