Generalized S transformation and SVM electric energy quality disturbance efficient identification method

Abstract

The invention relates to a generalized S transformation and SVM power quality disturbance efficient identification method comprising the following steps: collecting voltage data of a power quality disturbance signal, using a power quality monitor to collect the voltage data of the power quality disturbance signal, and transmitting the voltage data to an upper computer; gST generalized S transformation is carried out on the voltage data received by the upper computer, two groups of GST parameters are set, and a time amplitude envelope curve and a frequency amplitude envelope curve are respectively obtained; extracting a feature vector of the voltage data; and inputting the feature vector obtained in the previous step into a GWO-SVM classifier for training and testing, thereby completing identification of the power quality disturbance signal. The method has high identification precision in power quality disturbance identification, has strong anti-interference capability in power qualitydisturbance identification, and has good performance in processing a small sample disturbance identification problem.

Description

technical field [0001] The invention relates to the technical field of power systems, in particular to an efficient identification method for power quality disturbances of generalized S transform and SVM. Background technique [0002] With the development of smart grid, when various distributed power generation units are connected to the grid, various nonlinear loads are also increasing, causing a series of power quality problems. The identification of power quality disturbances is the basis and premise of solving power quality problems, so the efficient identification of power quality disturbance signals is of great significance. The identification process of power quality disturbance mainly includes two parts: signal detection and pattern recognition. [0003] At present, the commonly used signal detection methods include short-time Fourier transform, wavelet transform, Hilbert-Huang transform and S transform, etc. Among them, because the short-time Fourier transform use...

AI Technical Summary

Problems solved by technology

Among them, the neural network structure is simple, has strong solving ability and anti-noise ability, but there is a local optimal problem; the decision classification tree structure is simple, the classification accuracy is high, but the generalization ability is poor, and the optimal classification threshold is difficult to determine; extreme learning Machine parameters are easy to adjust and learning speed is fast, but the robustness and stability are poor; support vector machines have good performance in dealing with high-dimensional, small-sample, and nonlinear classification problems, but how to deal with penalty factors and kernel function parameters Reasonable setting is the core factor to improve the performance of SVM classification. As the training set increases, the number of kernel functions required increases significantly. It is sensitive to missing data and sensitive to the selection of parameters and kernel functions.

Images