Research Of Data Mining On TSA In Power System

As to the method of the security and stability of power systems, the experienced and qualified power engineers are necessary, and they have to find the weakness of power systems through tremendous calculation and decide the defensive strategy accordingly. However, the power systems become larger and larger, the quantity of the data captured by monitoring instrumentation has been increasing significantly, and the traditional method obviously can not match the great data capacity any more. Based on such tremendous raw data, how to extract useful information from them by Transient Stability Assessment(TSA) has become a critical issue, so as to determine the current operational state and predict some operational trend for operators effectively. In this paper, data mining technique is introduced for TSA to study the relationship between the operation schedule and transient stability when great quantity data involved.To the great operational data of power systems, a novel approach to assess the transient stability of power system, i.e. associative classification method(ACM) is given. Some associative rules reflecting the relationship between power system operating state and transient stability are generated and utilized to estimate the transient stability. This approach is independent from the mathematic method, so it can avoid the difficulty of parameters setting. ACM is based on frequency statistic of the operating samples, so some character features describing operating status can be included. For the results' uncertainty of the traditional ACM used in transient stability and the dynamic of power system operation, the improvement is gained in this paper. Due to lacking information about the dynamic characteristics of a power system, which is provided by the input features, misclassification is inevitable. So a fuzzy area is defined here by the aforesaid transient stability index and samples within it are assessed by Taylor series and time-domain simulation respectively according to their stability degree. Additively, time dimension is introduced into TSA. Some rules reflecting the effects the changes of influence factors making on the states changing of a power system can be obtained.