Power system transient stability assessment based on bayesian nonlinear hierarchical framework improved LSSVM

Abstract

This paper proposes an improved LSSVM power system transient stability evaluation method. According to the Bayesian nonlinear hierarchical framework, the automatic selection of parameters is implemented to improve the sensitivity of LSSVM classifier model to parameter and improve the accuracy of transient stability assessment Firstly, the original feature set related to the system stability is calculated by the wide area measurement information. The feature selection method is used to compute the feature, and the optimal feature set which is strongly related to the stability of the power system is determined. The training set and the test set are divided into Mapped to high-dimensional space, so that the nonlinear and linear classification of the transformation. Then, the optimal parameters of LSSVM are determined by Bayesian nonlinear hierarchical model, and the fast transient stability is determined. Finally, the validity and accuracy of the evaluation model are verified by IEEE-39 node system and actual system.