An evaluation framework for inverse hysteresis models

Control schemes employing inverse-based hysteresis compensation are successful in accurate position control of smart material actuators. However, the effectiveness of the inverse hysteresis model in cancelling hysteresis is not addressed in the evaluation of these control schemes. Classical methods rely on closed-loop tracking error analysis which does not allow evaluation of the inverse model independent of the controller. This leads to lack of a deterministic measure of the amount of hysteresis cancelled by the model, while the influence of an inaccurate model on the overall tracking error also remains unclear. This paper proposes a framework to verify the effectiveness and accuracy of inverse hysteresis models by quantifying the hysteresis non-linearities remaining after hysteresis compensation. Further, by estimating the linear and residual non linear dynamics, valuable information is provided for controller design. The framework is experimentally validated for a Shape Memory Alloy (SMA) actuator. The framework can be seen as a tool to explore and compare different inverse hysteresis models prior to controller design.