Observer and parameter estimation for IWP: Comparison of a super-twisting algorithm and an adaptive method

In this paper the comparison of an adaptive method with Super-Twisting for states and parameters estimation of an IWP is presented. The estimated states are arm velocity and pendulum wheel velocity through the position measurement (outputs) and the control signal (input). Subsequently the parameter estimator is implemented. The adaptive law estimates the states and parameters at the same time. For this, a passivity-based controller for positioning the pendulum in its inverted position is used. Finally the results simulated in Simulink / Matlab are presented.