Research On The Models And Approach Of Power Grid Security Assessment

Modern power system has gradually developed into a complex system with the features of multi-level structure, multi-time scale, uncertainties, non-autonomous and so on, so the security level of power system is difficult to be characterized by a single indicator. However, the conventional security analysis usually only focus on a certain aspect of power system security, and it is not enough to have a overall description of power system safety. Therefore, power system security assessment based on theory of comprehensive assessment is developed rapidly in recent years.Weight coefficient of index system is the kernel of the problem of comprehensive assessment. A two-level optimization programe which are optimization of indices and weight is presented. The model bring together the building of index system, quantitative index, index weight and index optimazition, Firstly, An index system for electric network security is presented base on the program and the feature of power grid. Secondly, Principal component analysis is used to optimiaze the system, which reduced the relevance of indices and simplified the process of evaluation. In addition, Specialist weight coefficient and weight coefficient of indices were analyzed from the point of subjective and objective conditions. Weight is obtained by FAHP which based on the credibility theory with the index weight of triangular fuzzy number in the process of subjective weight. To more subjective of weight, entropy weight is used to modify the weight which is obtained by FAHP to the variational index. In objective weight respect, firstly the tentative weight obtained by PAC, traditional comprehensive assessment method considers few about the balance of the indices, when there are great differences between the value of the indices, it is easy to take place the phenomenon that the small indices has been "submerged", which will influence the rationality of the comprehensive evaluation result. Considering the deficiency, objective weight is amended by introducing the Gini coefficient into the evaluation model. Since the coordinated and balanced situation of the indices have been reflected in the evaluation process. So the net result is obtained by subjective and objective optimization. The result of the method have a more reliable and mathematical theoretical basis. Finally an example was provided to verify the correctness and superiority of this method.