Research And Application Of Fractional-order Quantitative Feedback Theory

Quantitative feedback theory is an important branch of robust control theory. Compared with other robust control theories, it has the following several important characteristics:it can estimate the feedback cost and consider the phase information. It is independent of mathematical model, and has a wide range of application and strong engineering application, consequently a great success has been achieved in practical application. In robust control system, the design of controller is a vital step in the application of quantitative feedback theory, but the designed controller is usually integer-order controller. Along with the fractional-order calculus gradually penetrating the control field, fractional-order controller becomes the focus of research. Fractional-order controller can achieve more flexible control effect than integer-order controller, therefore the fractional-order controller design has become the focused research area.Previous QFT controller design is manually completed in QFT toolbox in matlab. This is a process of repeatedly trial and error, and can’t get the real-time feedback cost, therefore it is difficult to decide when the control performance is better. In recent years, automatic loop shaping controller design is a research focus. This thesis combines quantitative feedback theory with fractional-order controller. The main research contents include:Firstly, the paper introduces fractional calculus theory and basic knowledge of the fractional-order controller, including the definition of fractional-order calculus, different fractional-order controller forms and the realization of the fractional-order controller;Secondly, it introduces the basic principle and design process of quantitative feedback theory, and the roles of several important functions in QFT toolbox;Thirdly, it proposes optimization objective function more in line with the optimal QFT controller under the CRONE constraint, and optimizes by means of particle swarm intelligent optimization algorithm; By comparing with the designed different forms controllers, it proves that the controller using the constraint conditions proposed in this thesis has better robust performance and that fractional-order controller can achieve smaller feedback cost;Finally, the design of pre-filter is also an important part in quantitative feedback theory, but now few studies are related to the automatic design of the pre-filter. This thesis proposes the optimal pre-filter, implements using particle swarm intelligent optimization algorithm, and obtains the fractional-order pre-filter fulfilling performance requirements.