Development of robust stability and robust performance based control theory and applications

An open problem in disturbance/uncertainty estimator based control is to obtain explicit mathematical expressions for robust stability, performance, and bandwidth requirement. This problem can be divided into four categories as single-input-single-output (SISO) linear systems, multi-input-multi-output (MIMO) linear system, minimum/non-minimum phase systems, and nonlinear systems. In this thesis, a disturbance/uncertainty estimator based robust control approaches are studied to resolve this open problem respectively. Explicit mathematical expressions for robust stability, performance and bandwidth requirement are derived for all cases. The SISO and non-minimum phase parts of the theory are verified on a pan-tilt system and a rotary mechanical system. The MIMO part of the theory is verified on a full scale autopilot for a custom aircraft with significant couplings among its lateral and longitudinal channels. Moreover, an integral sliding mode controller is built and integrated into the robustness analysis via its quasi-linear representation. The proposed methodology is experimentally verified on a high-precision gimbal control application. Comparisons with state-of-the art methods in literature show noticeable performance and robustness improvements.