Nonlinear MR model inversion for semi-active control enhancement

Abstract

The increased prevalence of semi-active control systems is largely due to the emergence of cost effective commercially available controllable damper technology such as Magneto-Rheological (MR) devices. Unfortunately, MR dampers are highly nonlinear, which presents an often over-looked complexity to the control system designer. The well-known Skyhook Damping control algorithm has enjoyed great success for both fully active and semi-active control problems. The Skyhook design strategy is to create a control force that emulates what a passive linear damper would create when connected to an inertial reference frame. Skyhook control is device independent since it generates a desired control force command output that must be produced by the control system. For simplicity, MR dampers are often assumed to have a linear relationship between the current input and the force output at a given relative velocity. Often this assumption is made implicitly and without knowledge of the underlying nonlinearity. In this paper, we show that the overall performance of a semi-active Skyhook control system can be improved by explicitly inverting the nonlinear relationship between input current and output force. The proposed modification will work with any semi-active control algorithm, such as Skyhook, to insure that the controller performance is at least as good as the performance without the proposed modification. This technique is demonstrated through simulation on a quarter-vehicle system.

A novel model of magnetorheological damper with hysteresis division

“…The phenomenological models rely largely on the shape functions to describe the non-linear hysteretic characteristics. The existing shape functions contain the differential equations [18][19][20], trigonometric-family functions [21,22], exponential-family functions [23,24], piecewise functions [1,25,26] and so on [27][28][29][30].…”