Nonlinear and adaptive control with applications to active suspensions

The focus of this dissertation is nonlinear and adaptive control with applications to an automotive active suspension. Within the scope of this research, the control of an active suspension is defined as the ability to track a desired force profile. The goal is the reduction of sprung mass acceleration for increasing the ride quality. The suspension contains an electro-hydraulic actuator which possesses nonlinear characteristics. The dynamics of the actuator are developed using standard Fluid Mechanics and are coupled with the dynamics of the suspension system which are acquired from standard Newtonian Dynamics.;A sliding controller is designed for the suspension system using a two surface approach. The two-surface approach is used to separate the problem into two simpler problems. Stability and convergence issues are discussed and proven. The sliding controller relies on an accurate model of the system dynamics in order to avoid large control gains. To reduce the error in the model, a standard parameter adaptation scheme is applied. The adaptation algorithm is developed by a Lyapunov analysis. The parameter estimation method is then modified to exploit the dynamics of the system. The modification enables the identification of a class of parameters whose values change with regions in the state-space. Therefore, the parameters are not restricted to being constant or slowly time-varying and can vary as quickly as the states themselves; however, the parameters are restricted to being slowly time-varying within sectors of the state-space. The modified algorithm is coupled with the sliding control and the stability and convergence of both the output error and the parameter error are analyzed.;The controllers are implemented in simulation and experiment. In both cases, the active suspension system performs better than the passive suspension in reducing the acceleration of the sprung mass. The adaptive controllers perform better than the fixed-parameter controllers with the modified adaptive scheme performing better than the standard approach.