Analysis And Synthesis Of Complex Systems Using Gain Scheduling

Facing the shortage of the conventional gain scheduling control, this thesis focuses on the approach of gain scheduled controller designing for nonlinear systems, Markov neutral systems with time-varying parameters or LPV systems, etc. Fuzzy H_∞controller, robust observer, dynamical output controller and L₂ -L_∞filter that with continuous gain scheduled parameters are respectively designed.Generally speaking, the first step in the controller design procedure is to obtain a mathematical model of the dynamic process of systems. Then, analysis this model and design the controller. However, the truth model is nonlinear or with time-varying parameters, so it will bring some difficulties in the controller design. The proposal of gain scheduling control provides a new approach of dealing with the controlling problem for complex systems which don't have precise mathematical models.The major contributions of this thesis are as follows:1 For the systems with uncertainties such as hyperchaotic systems, CSTR systems, LPV systems, etc. the grads linear approach is invited to obtain the sectional linear models at typical operating regions and time regions. Then H_∞feedback controller, Min-Max optimal feedback controller and Min-Max optimal output feedback controller are designed respectively. Lastly, the fitting approach of Taylor series is investigated to obtain the continuous gain scheduled controllers.2 For the systems with actuator saturation and subject to uncertainty and time-delay, continuous gain scheduled H_∞observe is designed.3 For the Markov neutral systems with time-varying parameters, continuous gain scheduled H_∞feedback controller and L2 -L∞filter are designed respectively based on the exponential stability theory.