Estimation of and compensation for unknown input nonlinearities using equivalent-input-disturbance approach

Abstract

Nonlinearities greatly affect the control performance of a system. A new estimation and compensation method for unknown input dead zone and backlash has been developed based on the equivalent-input-disturbance (EID) approach. Unlike other methods, it requires neither any information on the nonlinearities nor an inverse model of them. The key idea is that input nonlinearities are taken to be an input-dependent disturbance. An EID estimator is designed that produces an estimate with the same effect on the output as that of the input nonlinearities. The estimate is then used to compensate for the nonlinearities. Simulation results demonstrate the validity of the method.