Polynomial linearization for real-time identification of environment Hunt-Crossley models

Abstract

Mathematical models describing physical environments are often used in robotic and haptic systems. The Hunt-Crossley (HC) model has been shown to be more accurate and physically consistent than the Kelvin-Voigt model for deformable environments such as soft tissues. In this paper, a novel real-time identification method is presented in which a linearly parameterized polynomial approximation is used to indirectly identify the HC model. Simulation results show that "Polynomial Identification" excels in highly damped environments, which presented a challenge in previous methods. The method is also shown to be less sensitive to estimation parameters and more robust during intermittent contact.