Signal Prediction in Bilateral Teleoperation with Force-Feedback

In the paper a sensor-less and self-sensing control scheme for a bilateral teleoperation system with force-feedback based on a prediction of an input of a non-linear inverse model by prediction blocks was presented. As a part of the paper a method of a time constant estimation of the prediction block was also proposed. The prediction method of an input of an inverse model was designed to minimize the effect of the transport delay and the phase shift of sensors, actuators and mechanical objects. The solution is an alternative to complex non-linear models like NARX or artificial neural networks, which requires complex stability analysis, and control systems with high computing powers. The effectiveness of the method has been verified on the hydraulic manipulator's test stand.