Prediction-Based Adaptive Control of Haptic Devices Under Communication Time Delay and Human Hand Interference

Abstract

This paper presents a predictor-based adaptive con trol framework to improve the performance of haptic interaction in the presence of time delay, sampling effects, and uncertain human hand dynamics. In haptic systems, computational delays, zero-order hold effects, and variations in hand impedance can de grade stability and prevent the device from accurately following the continuous trajectory generated by the virtual environment. To address these challenges, the predictor is designed to generate a delay-free trajectory for the adaptive controller, while the adap tive controller estimates uncertain hand impedance parameters and compensates for non-passive forces exerted by the user in real time. To reduce implementation complexity, the effects of sampling and zero-order hold are represented using Taylor series approximations, leading to a more practical theorem for predictor design. The stability of the proposed predictor is established using the linear matrix inequality (LMI) method, and its performance is evaluated in MATLAB/Simulink by comparing it with the widely used Smith predictor.