LPV Control Approaches in View of Comfort Improvement of Automotive Suspensions Equipped with MR Dampers
Résumé
Many studies have shown the importance of automotive suspension systems in vehicle dynamics, see for instance [10], [26], [33] and references therein. Except for passive suspensions whose characteristics are invariant, the semi-active and active suspensions can change their properties by using controlled external signals (voltage, current...). This is why the latter suspensions have been studied intensively in recent years. However, up to now, only the semi-active suspensions are used widely in automotive industry. Indeed, comparedwith fully active suspensions, semi-active ones can achieve the main performance objectives (see [17], [27]) while they are smaller in weight and volume, cheaper in price, more robust and less energy consuming (see also [9], [10], [16], [19]). So far, the control problem for semi-active suspensions has been tackled with many approaches during the last three decades. One of the first comfort-oriented control methods, successfully applied in commercial vehicles, is the Skyhook control proposed by Karnopp et al. [18]. Then, optimal control [12], [34], clipped optimal control [24], [36], [11], H∞ control [30], [31] or Model Predictive Control [4], [28] have been considered. Recently, two new control design methods for semiactive suspensions using the LPV techniques have been presented. The first one, proposed in [29], can be applied for all kinds of semi-active dampers where only the bounds on damping coefficients and on the damper forces are necessary for the controller design. In the other one, proposed in [7], the nonlinearities of the semi-active damper (the bi-viscosity and the hysteresis) are taken into consideration. The comparison of these two recent LPV-based techniques on a nonlinear Magneto-Rheological (MR) damper model is proposed this chapter. The chapter is organized as follows: In section I, a brief bibliography concerning the modelling of semi-active dampers is given and two specific control-oriented models are detailed and will be used for the synthesis of the LPV controllers. In section II, the control problem of automotive suspension control is formulated in a common way so that the methods proposed in [29] (section III) and [7] (section IV) can be applied. Section V is devoted to numerical simulations on a nonlinear quarter car model. Some remarks and conclusions end this chapter.