| This dissertation consists of three aspects of the research efforts in control system designs: 1. An observer for nonlinear control systems with partial input-output feedback linearization, 2. System modeling and parameter identification, 3. Applications of the observer-based nonlinear controls to biomedical and automotive engineering systems.;In the first aspect, the problem of the observer formulations for a control-affine nonlinear system with partial input-output feedback linearization is considered. When a nonlinear system is partially linearizable -- the relative degree is less than the dimension of the system -- and resulting in a normal form, it is well-known that its internal dynamics are unobservable by the output, and a full-order observer is impossible to be formulated. This research focuses on the manipulations of the external and internal dynamics of the normal form in such a way that the augmented system, which consists of the external and internal dynamics, yields desirable properties that enable a full-order observer to be constructed. The research results suggest a procedure and the sufficient conditions that a full-order observer for the normal form can be formulated.;In the second aspect, the system modeling and identifications of an electronic throttle control system are considered. This research fulfills an important aspect of control system designs in which a mathematical model is a necessity. It also serves well from the practical view points as the identification process involves extensive laboratory experiments, measurements, and data processing techniques.;In the third aspect, two application systems in two different areas are investigated: the heartbeat control systems in the biomedical engineering field, and the active suspension system incorporating a negative stiffness spring in the automotive engineering area. The open-loop stability analysis of both systems is conducted. The control design for both systems is based on the input-output feedback linearization methodology. The proposed observer technique, which is the research results in the first aspect, is applied successfully in the heartbeat systems while the Luenberger observer is applied to the active suspension system. Simulations are conducted for both applications and results are reported. |
