Application of Nonlinear Estimation Strategies on a Magnetorheological Suspension System with Skyhook Control

Abstract

Extraction of state values from noisy or uncertain systems is important for feedback control because it improves the accuracy of the error signal. For known linear systems with Gaussian white noise, the Kalman Alter provides optimal state estimates in terms of state error. However, electromechanical systems, such as magnetorheological dampers, typically exhibit nonlinear behaviour. In this paper, a new nonlinear estimation method known as the extended sliding innovation filter is presented and applied on a magnetorheological suspension system. The state estimates are extracted from a quarter car model with an active magnetorheological suspension system with skyhook control. The results are compared with the popular extended Kalman filter, and future experiments are considered.