The Study Of Power System Model Order Reduction Method Based On The Theory Of Proper Orthogonal Decomposition

Power system mathematical model is a data file used to describe the system global running state, and it is also the foundation of the grid services. With the expanding scale and the increasing complexity of power system, the dimension of system mathematical model is also increasing rapidly. At the same time, the computing time also increases obviously. It directly leads to the problems of the power system controller design and the simulation analysis becoming more and more prominent. In consequence, the research of how to simplify the system model and reduce the dimension of system model, has an important academic meaning and an engineering application value. Based on POD(Proper Orthogonal Decomposition) theory, this article studies the model reduction method of power system. The main content includes the following aspects:(1) The research status of all kinds of order reduction methods at home and abroad is introduced. At the same time, the advantages and the disadvantages of these various order reduction methods are analyzed and compared. Then, the power system model order reduction method based on POD theory is put forward.(2) In view of the general linear and nonlinear system mathematical model, the basic principle of POD algorithm is respectively introduced. At the same time, the order reduction process of power system mathematical model is analyzed. Through the acquisition of system sample vector set, the singular values of vector set in descending order and the corresponding right singular value vectors can be obtained. Then, according to the size distribution of these singular values, the main state dimension of the system can be determined. Next, by Galerkin mapping, the corresponding state of these important singular values can be projected onto the subspace and the reduced-order model can be obtained. In addition, the error estimation methods and the decision basises of stability for linear and nonlinear reduced-order models are respectively provided.(3) The linear-model building method of power system is studied. Then the linear model, which is applied to the model order reduction of power system, can be obtained. Taking the single machine-infinite bus system and the nine order area power system, for example, the programming for the order reduction process of power system linear model is achieved. At the same time, the simulation analysis results for linear reduced order model of power system are given. In addition, the availability that the POD order reduction method can be applied to the linear-model order reduction in power system is verified.(4) The nonlinear-model building method of power system is studied:through the conversion of coordinates, the interfaces which include the generator model, the network model and the load model, etc, can be handled and then the nonlinear model can be obtained. Taking the single machine-infinite bus system and the three machines nine nodes system, for example, the programming for the order reduction process of power system nonlinear mode is achieved. At the same time, the simulation analysis results for nonlinear reduced order model of power system are given. In addition, the availability that the POD order reduction method can be applied to the nonlinear-model order reduction in power system is verified.The innovation significance of the model order reduction method proposed to apply to power system in this paper is mainly embedded in the following three aspects. Firstly, the method is suitable for both linear and nonlinear model order reduction of power system. Secondly, compared with the traditional order reduction method, the calculating process of this method is much simpler. Thirdly, the order reduction algorithm has the general applicability, which has a wide range of application.