Development of a nonlinear adaptive control scheme for hydraulic actuators

Abstract

Hydraulic servovalve controlled systems contain many time-varying dynamic characteristics that are difficult to model. Controllers for such systems must either adapt to these changing parameters or be robust enough to handle the parameter variations. In this Thesis the application of an indirect model reference adaptive control (MRAC) scheme to a hydraulic actuator is investigated. A discrete-time plant model is suggested which has dead time and no zeros, eliminating the possibility of unstable pole-zero cancellation. Using this model, a linear indirect adaptive controller is developed and tested first on a simulation model of a hydraulic positioning system The robustness of the parameter adaptation is achieved by using recursive least-squares algorithm in combination with a dead zone in the adaptive law. While the controller shows good performance for a valve with linear flow characteristics, it shows low performance in the presence of flow deadbeat and nonlinear orifice opening. To deal with these two nonconformities, the linear plant model is augmented by adding a static nonagenarian. The resulting nonlinear MRAC shows improved performance as compared to its linear counterpart. Experiments on an existing hydraulic test station confirms the effectiveness and performance improvement of the proposed method.