Research On The Time Delay Method Of Internet-based Teleoperation Robot System

The invention of teleoperation robot would be aimed at assisting human to completecomplex tasks in harmful or inaccessible environment, meanwhile creating more socialvalue. Teleoperation robot can also be applied to explore the space and the deep sea, providerescue after natural disasters (e.g. earthquake), operate in hazardous environment (e.g.nuclear), etc, which greatly expand the space of human activities.With the rapid improvements of computer network, Internet-based teleoperation robotsystem has emerged. However, the presence and random of network time delay, packets lossand disorder seriously impact on the performance of the teleoperation robot system, whichinclude stability and operability.Considering the impact of the time delay to Internet-based teleoperation robot system,the main contributions of the paper were as follows:a) It expounded on the development of Internet-based teleoperation robot system andthe research summary of time delay control algorithm. Furthermore, it analysed the forcefeedback teleoperation robot system performance indexs (stability and transparency) basedon the2-porst network model with the passivity and scattering theory, meanwhile the impactof network time delay to the performance of teleoperation robot system.b) It established the Internet-based teleoperation robot system time delay controlalgorithm simulation platform. Due to the problem that traditional simple single closed-loopcontrol structure could not satisfy the complex characteristics of Internet time delay, thesimulation platform had been established based on the force deviation feedback bilateralcontrol structure. In the simulation platform, the plant model of three degrees of freedom(3-DOF) teleoperation robot had been established with mechanism simulation toolboxSimMechanics; the communication node had been established with network control systemsimulation toolbox TrueTime; the controller node had also been designed by the TrueTimemodules, users could program different controller codes in order to realize correspondingcontrol algorithms, which made the time delay control algorithm simulation more flexible.c) It designed controller based on the conventional PID control algorithm. Consider theproblem that conventional PID controller is unable to control time delay and its tuningparameters methods are complicated when been used in the nonlinear and time-varyingpractical applications, it presented the fuzzy PID control algorithm. Meanwhile on the basisof the standard BP neural network control algorithm theoretic research, it had find theexisted shortcomings of the algorithm, such as adopting the mean square error gradientalgorithm could result in falling into local minimum easily and reducing the convergencespeed, consequently it presented the improved BP neural network PID control algorithm.d) It carried out the simulation on the designed controllers based on Internet-based teleoperation robot system time delay control algorithm simulation platform, and analyzedthe simulation results. Intelligent control algorithm could self-tune the PID controlparameters, consequently it had better time delay control effect than conventional PIDcontrol algorithm. The results of the simulation shown that the improved BP neural networkPID control algorithm had better time delay control effect than fuzzy PID control algorithm,and it was optimal to the time delay control of Internet-based teleoperation robot system.