A method to recognize transient voltage disturbance in power system

It is difficult to recognize the transient voltage disturbance in power system, because of the low-energy and short duration of transient voltage signal. To solve this problem, a data mining method, by the combination of Tsallis entropy and wavelet transform, is proposed firstly. After analyzing the statistic characteristics of Tsallis entropy, Tsallis wavelet energy entropy (TWEE), is put forward with wavelet transform, and applied to extract the feature of voltage signals corresponding to transient disturbances in power system. Based upon TWEE and clustering algorithm, a novel method for the recognition of transient voltage disturbance was presented and applied to classify the different transient disturbances in power distribution system. From the feature of transient voltage by TWEE, the feature criterions of transient voltage disturbance are summarized to constitute classification rules with clustering algorithm. Through arranging the feature criterions, the optimized classification rules were obtained. By use of Matlab/Simulink, a simulation model of transient voltage disturbance was built to verify the proposed method. Simulation results showed that this method can simplify the recognition process of the transient voltage disturbance and improve the classification accuracy.