Classification and identification method of transient power quality disturbance based on cwd spectrum kurtosis

Abstract

The invention provides a disturbance classification and recognition method for the quality of transient electric energy based on CWD (Choi-Williams Distribution) spectral kurtosis. The disturbance classification and recognition method is combined with effective values for recognizing the disturbance of five types of single transient electric energy and complex transient electric energy. The disturbance classification and recognition method comprises the following steps of: firstly taking three types of amplitude actions, i.e. voltage swell, voltage sag and voltage interruption as the same class, classifying data acquired by a data acquisition system of an electricity bureau into transient pulse, transient oscillation and amplitude disturbance through CWD spectral kurtosis values, then classifying the data into the voltage swell, voltage sage and voltage interruption by calculating the effective values of the amplitude disturbance, and outputting obtained classification results to a follow-up processing device. According to the disturbance classification and recognition method, different disturbance types are distinguished by directly using the size of values, no classifier is needed, and the recognition flow and time are greatly simplified. The disturbance classification and recognition method can be used for accurately distinguishing the five types of single transient disturbance, i.e. transient pulse, transient oscillation, voltage swell, voltage slag and voltage interruption, and the complex disturbance formed by combining the five types of single transient disturbance; and according to the method, good noise immunity is achieved.

Description

technical field [0001] The invention relates to the intelligent monitoring of power systems, in particular to the technical field of classification and identification of transient power quality based on high-order statistics and signal processing. Background technique [0002] With the continuous expansion of the scale of the power system and the massive investment of various power electronic equipment, nonlinear loads, and impact loads, various disturbance events in the power system have seriously affected the quality of industrial products and daily life. Since different types of power quality disturbances have different influence degrees, it is very important to identify power quality disturbances. [0003] The recognition of power quality disturbance includes two steps of feature extraction and classification recognition. Commonly used feature extraction methods include short-time Fourier transform, wavelet transform, S transform, HHT transform, etc. Among them, the mea...

AI Technical Summary

Problems solved by technology

The training speed of the conventional ANN classifier is slow, it cannot provide the local details of the signal, and its accuracy needs to be improved; the SVM classifier has a short training time, high recognition accuracy, and is not sensitive to noise, but this method is effective in identifying mixed disturbances. It is more difficult; ES classifiers are prone to combinatorial explosion problems when the types of power quality events increase

Moreover, the above classifiers require a large amount of data for training and testing, and the actual power quality data is not easy to obtain, which makes the algorithm difficult to implement in practice

Images