Compensation for time delay in tele-operation control of mobile robot using neural network

Abstract

In communication channel of a tele-operation control task, there always exists a time-delay problem which causes poor performance and even instability of the system. To overcome the time-delay problem, in this paper, the Smith predictor configuration is considered. The Smith predictor can stabilize the system by taking the time delay term out of the characteristic equation when the system model is given and a delayed time is known. Neural network is employed to identify and model the slave controlled system to form the structure of the Smith predictor. Experimental studies are conducted to regulate contact force of a mobile robot by the master robot, the Phantom haptic device.