Identification Of Uncertain Systems Based On Robust H_∞ Filter

The traditional methods of system identification identify the unknown parameters by the input and out signal based on the assumption that some system parameters which are known are certain. While in the practical application, most systems have uncertain parameters and uncertain perturbation from the outside, and if we still use the traditional methods, it will make the identification abortive. So we should search a robust identification method which can deal with the uncertain systems.In this paper, the robust H_∞ filtering identification problem for discrete uncertain system is studied, and the main research work and achievements are as follows:(1) First, in this paper the filtering identification theory is systematically introduced, especially the development process and research actualities of the filtering identification theory. The concept of filtering identification and the typical algorithms, conclusions which come from H₂ optimal filtering identification, H_∞ filtering identification, and robust filtering identification are presented briefly.(2) Second, the robust H_∞ filtering identification problem for discrete uncertain system is studied in this paper. I use a game theory approach that incorporates maximum principle arguments to give a new robust H_∞ filter.(3) Use the single degree of freedom and three degrees of freedom system to confirm the new robust H_∞ filter when the system is certain and uncertain respectively, and compare the estimation error between the new robust H_∞ filter with and the standard H_∞ filter.The considerations and conclusions of the research are given in the end of the paper. Meanwhile, a view is given on the prospect of the filtering identification theory and the problems that the filtering theory may meet are stated as well.