Gain Scheduled And Its Application On The Objects In Thermal System

Thermal objects are usually nonlinear systems whose parameters vary in a large region. In this dissertation, some control schemes based on gain scheduled and state feedback control theory and PID controller were proposed to control the nonlinear parameter-varying thermal objects. In present dissertation, through the stability analysis with the state feedback control, Kharitonov theory, and generalized Hermite-Biehler theorem, gain scheduled was studied for thermal control problems, which were confirmed by the computer simulations. Firstly, a stability-preserving interpolation method was proposed to compute the controller parameters in the operation regions between those equilibrium points. It averted the problem of real-time computation for two matrix inequations and interpolation with matrixes, which the classical gain scheduled interpolation method had to meet. This method was implemented in controlling a three-input and three-output nonlinear boiler-turbine unit with parameters varying in a wide range. Secondly, bumpless switching methods were proposed to extend gain scheduled to those plants which used different form controllers in the operation region. These methods were implemented in global control for the angle and distance of the inverted pendulum. Thirdly, based on Kharitonov theory, stability-preserving gain-scheduled PI controllers were designed for a kind of two-order plants whose parameters varied in a large range. This method was applied in a separate-axis gas turbine for automobile use with operation points varying globally. In addition, this method provided a scheme to combine gain scheduled and PI controllers. Finally, based on simulation study, a control scheme for the nonlinear unknown system, especially for thermal objects, was provided. It included a system identification method proposed to identify the order and parameters of an unknown system and PID-controller bumpless switching method which was based on the generalized Hermite-Biehler theorem to guarantee the system stability. This control scheme was confirmed to control a simulation model of the nuclear steam generator water level. In addition, this control scheme provided design experience for gain scheduled to be implemented in industry, especially in thermal object control. The study of this dissertation illusrtated the promising future of gain scheduled in thermal object control.