Author(s)

Li Xin, Fan Youping, Zhang Peng, Liu Songkai

Corresponding Author

Li Xin

Abstract

A general probability evaluation method is proposed to evaluate the accuracy of the data mining model in the on-line prediction of transient stability after the failure. Considering the distribution of failure probability, the paper first searches for possible faults, then evaluates the accuracy according to the actual probability distribution of uncertain factors, which is more objective than Monte Carlo method. Finally, simulation is carried out with the new England 39 bus system, the test results show that this method can not only comprehensively evaluate of data mining classification decision model to take timely emergency measures, but also compare the mining accuracy models in the prediction of the objective of different data.

Keywords

Transient stability assessment, support vector machine, Probability evaluation, on-line prediction.