Prescribed performance model-free sliding mode control using time-delay estimation and adaptive technique applied to industrial robot arms

Abstract

This paper introduces a novel prescribed performance model-free controller tailored for industrial robot arms, seamlessly integrating adaptive sliding mode control (ASMC) and time-delay estimation (TDE). Leveraging TDE, our controller adeptly estimates both the inherent dynamics of the robot and unstructured uncertainties such as disturbances and parameter variations. However, TDE, which relies on past angular acceleration and input torque, inevitably introduces errors. To mitigate these, our approach compensates for current TDE errors using past error information. Additionally, we introduce a fixed-time sliding mode surface from prescribed performance control and an auxiliary system to improve performance under input saturation. Moreover, we propose an adaptive law to ensure the positivity of the adaptive parameter by considering the current adaptive parameter value and the sampling period. Through extensive simulated studies conducted on industrial robot arms, we demonstrate the effectiveness of our control approach, showcasing robustness, reduced chattering, and high accuracy across diverse scenarios.