A fuzzy-based dynamic inversion controller with application to vibration control of vehicle suspension system subjected to uncertainties

Uncertainty or the model error always exists in depicting technique systems. In many cases, the model error is implicitly the time-varying tendency such as in the semi-active railway vehicle suspensions using magneto-rheological damper, where the hysteretic response and time varying of physical parameter value of the magneto-rheological fluid due to system temperature variation are inherent features. To exploit well these systems, a becomingly compensative and adaptive mechanism needs to be set up to deal with both uncertainty and external disturbance. In this study, a new fuzzy-based dynamic inversion controller for semi-active railway vehicle suspensions using magneto-rheological dampers subjected to uncertainty and external disturbance is proposed. The fuzzy-based dynamic inversion controller faces uncertainty and disturbance individually. Regarding external disturbance, a disturbance observer is designed to estimate the compensative quantity. While for the model error influence, a combination of the inference ability of an adaptive fuzzy system and the competence to reach and keep the system stability states of sliding mode control is established via a model so-called fuzzy sliding mode control. First, an optimal sliding mode controller is designed, which is then used as a framework for building the adaptive fuzzy system. Based on Lyaponov theory, adaptive update law for the adaptive fuzzy system is discovered to adjust the sliding mode control adaptive to the system status. Surveys via quarter and half train-car vehicle suspension models including a real system of semi-active railway vehicle suspensions using magneto-rheological damper have been performed, which reflected the positive ability of the proposed method to stamp out vibration.