Research On Robust Control With Region Pole Based On Uncertain Sampled-data Systems

In the actual industrial control systems,due to the existence of various uncertainties such as model errors and unknown disturbances,robust control theory is proposed to study the asymptotic stability and other performances of the closed-loop system.Most of analysises on robust control can be transformed to the standard form of robust H_∞control theory.Considering the shortcomings that parameters adjusted manually in the process of solving the Riccati equation is fussy,the method of the linear matrix inequality(LMI) is adopted in the thesis.LMI can convert the H_∞control problem to the combinatorial LMI convex optimization problems.The result of corresponding controller can be easily solved by LMI toolbox in the MATLAB.Although the problem of system's robust performance can be better solved by H_∞control theory,it costs other performances of systems.The thesis focuses on the robust control method with the regional pole placement that the circle center is q,the radius is r,which can satisfy the robust stable performance and disturbance restrained performance.The research works are as follows: Firstly,after briefly reviewing the development of the robust control theory and the regional stability,a class of state feedback controller is designed with the method of LMI,which can make the closed-loop sampled-data system satisfy the robust H_∞performance with regional pole constrain.Then the controller with smaller gain parameters can be obtained by the method of D stability.At last,a simulation example is given to demonstrate the conclusion that the more the pole in the right half of the unit circle close to the spot,the better the dynamic performance of system.Secondly,in practical control systems,it's necessary to use observer for state informations are unable or uneasy to be observed because of noises.A class of sufficient conditions of the robust controller of the sampled-data systems are given via solving a class of LMI constrains.The controller and state observer not only reconstructs the state information , but also improves the system performances.Then a numerical example with the simulation is given to show the correctness of present sufficient condition and effectiveness of present method.Finally,the observer results of conventional state observers have big errors when unknown disturbances exist.Based on the properties of the proportion link and the integration link,error integral of output between the pratical system and the observer system is used.Proportional-integral (PI) observer is designed to enhance the ability of sustaining disturbance with observer.A class of sufficient conditions satisfied the PI observer with the region pole placement are obtained via the method of LMI.Compared with the previous examples,PI observer has better ability to reconstruct the state , much better disturbance-restrained performance and improved dynamic performances.