Dynamic optimization of nonlinear semi-active suspension controllers

Active and semi-active suspensions offer potentially significant performance improvements over conventional passive suspension. Several authors have applied linear optimal control theory to active suspension design; however formal optimization of semi-active controllers is much less simple, and few have applied rigorous methods in this context. The standard technique of 'clipping' a simple linear control law-applying the closest available suspension force appears to yield acceptable results. Here we shall consider a reasonably realistic suspension ride model, incorporating actuator dynamics and damper compliance. The formal optimization procedure involves extending previous nonlinear optimal designs to the semiactive system. Even for an ideal active system, there are performance benefits available, compared to linear optimal controllers. There are a number of factors that complicate any optimization: nonlinear coupling via damper rate control; limited actuator bandwidth; compliance in the damper operation; nonlinear damper characteristics; and limited maximum and minimum damper rates. The aim of this paper is to compare performance of a number of semi-active control laws, restricting attention to simple initial condition disturbances. In particular, the effectiveness of clipped linear controllers is considered.