Modelin And Nonlinear Control For DC/DC Switching Converter

DC/DC switching converters are time-variable nonlinear systems. The dominant approaches employed in stability analysis and controller design for DC/DC switching converters are based on the state-space averaging models, which are a result of simplifications that only include averaging behavior, i.e., the slow states of the system. Thus merely under small perturbation condition can the derived controller perform well. In this thesis, modeling and control techniques are explored in an aim of applying theoretical results of nonlinear system to improve the performance of DC/DC converters.On the basis of the current situation and developing trend of modeling methods of DC/DC switching converters discussed, firstly, from an engineering point of view, a method of PID parameter optimization based on linear quadratic optimal control is shown in this dissertation. With this method, the controller of the PWM DC/DC converter is designed. Utilize augmented matrix to build a tracking control model of DC/DC switching converters with no steady error, and through select weighted matrix to perform the no overshoot and dynamic response of system.Secondly, this dissertation regards the small-signal modeling of DC/DC converters as a class of uncertain parameter systems. Considering the disturbance of input voltage and resistive load, a novel robust-optimal control is proposed. The result proved that the method can effectively restrain disturbance and system has desired property of stability robustness and tracking performance. Finally, from the combined perspective of switched system and passivity system theory, a novel approach to the modeling of DC/DC converters is presented, namely switched affine systems. Then, based on the sufficient condition for quadratic stability of convex combination of the controllers, together with used the stored energy in the inductance and the capacitor as a common Lyapunov function, switching rules and activation region of all the subsystems can be defined to assess quadratic stability of the switched system.