Construct a Real-time Selective Neural Group Network Controller for Internet-Based Teleoperation System

The existence of random time-delay in the Internet makes telerobot unable to receive the timely correct instructions. This may seriously decrease the stability and reliability of the system. To solve this problem, many researchers have turned their interest to increasing the local-autonomy of the telerobot. In this paper, a real-time selective neural group network controller for the local telerobot system is presented and described in details. In the simulation experiments, we compared it with the traditional fuzzy logic strategy. Simulation results showed: the ability of on- line learning, information accessing and the predicted evaluation about selective strategy are much higher than the fuzzy logic strategy.