Research On Predictive Control Of Tele-operation Robot With Network Delay

Internet-based telerobots are widely applied to nuclear stations,space,subsea exploration,surgical operations and remote education,etc.However,there are time delays which can degrade the performance of Internet-based telerobot,even make the system unstable.The time delays in the Internet-based telerobot are random and uncertain.In order to slove the problem of time delays,an architecture based on improved generalized predictive control with predictive time delays and PID control is proposed.Firstly,the system structure of the Internet-based telerobot is analyzed.The dynamics equations are established based on Lagrange function.Thus,the state space models of the system can be obtained.The theory of predictive control and grey model are introduced in this paper which lay the foundation for the design of prediction controller and grey predictor.Then the measured time delays are analyzed to establish models.By comparison with other predictive models,grey model can get precise predictions with less amount of modeling data.The results show that the grey model is more suitable for rolling forecasts online.Based on the grey prediction of time delays,the system architecture and implementation methods are given.The master controller JGPC calculates a number of control prediction values and the slave is close-loop in which PID can use the predictions to realize periodic control.At last,by analyzing and comparing the simulation results with bilateral PID based on wave variable,it demonstrates that the proposed method can make the system with varying time delays get faster response speed and guarantee its stability and control accuracy,at the same time.