Characterization of Semi-active Control System Dynamics with Magneto-rheological Suspensions

Abstract

This paper characterizes two semi-active control algorithms through strategy analysis and parametric simulation study. A brief literature review discloses two kinds of semi-active controls: non-model based skyhook-like control policy (or skyhook variations), and model-based clipped semi-active control strategies. Both can be formulated as a bi-state control algorithm. In this paper, the no-jerk skyhook control policy is chosen to represent bi-state strategies, while a nonlinear adaptive semi-active control algorithm without need of clipping control signals is selected for comparison study. Then the studied MR suspension and its model setup are presented. In the paper we explain dynamic analysis approach and controller tuning for this simulation study. While analytic analysis of strategies is based on linear base-excited vibration systems by using transmissibility plots, simulation study adopts sine-sweep excitations so that comparison of transient dynamics could be stressed out and simultaneously the vibration isolation capability compared in the frequency domain by directly observing the time domain data. In addition, one pure-tone excitation case is simulated to explain higher harmonics inherent with bi-state skyhook controls. From the simulation study, the effectiveness of both semi-active controls can be observed for vibration isolation. In addition, the adaptability of both selected controls is compared regarding plant variations and different excitation profiles, both of which are typical in vehicle suspension systems. Finally the characteristics of these two semi-active controls are summarized from the simulation findings.