基于S变换与多维分形的电压暂降源特征识别

Voltage Sag Source Feature Identification With S Transform and Multidimensional Fractal

电压暂降源的分类与识别是合理制定电压暂降治理方案,明确事故责任的基础。分析了由系统短路故障、大型感应电动机启动和大容量变压器投运引起的单一暂降信号和复合暂降信号的特征,采用S变换分析暂降信号的基频幅值变化情况,对变换后的模矩阵提取6种特征指标,提出多重分形谱参数广义Hurst指数提高噪声环境下分类识别的准确性,将两者结合共同构成电压暂降信号的特征指标。将提取的特征指标作为支持向量机的输入,对不同类型的电压暂降进行训练,并利用无噪声数据和仿真加噪数据分别对其进行测试,从而实现对不同暂降源的分类与识别。实验结果表明,对比传统S变换的电压暂降源识别方法,采用S变换和多重分形相结合的方法构建的特征指标可以更好地识别电压暂降源,有效改善含噪声信号的分类效果,能够应用于实际工程。

The classification and identification of the voltage sag source is a reasonable basis for formulating the voltage sag treatment and clarifying the possible accident liability. The characteristics of single sag signal and composite sag signal caused by system short-circuit fault, large induction motor start-up and large-capacity transformer commissioning are analyzed. The S-transform is used to analyze the change of the fundamental frequency amplitude of the sag signal. The modular matrix extracts six characteristic indicators, and a multi-fractal spectral parameter generalized Hurst index is proposed to improve the accuracy of classification and recognition in a noisy environment. Combining S transform and multi-dimensional fractal together constitutes the characteristic index of voltage sag signal. The extracted feature indexes are put into the support vector machine, and different types of voltage sag are trained, and the noiseless data and the simulated noise-added data are respectively tested to realize the classification and recognition of different sag sources. The experimental results show that compared with the traditional S-transformed voltage sag source identification method, the characteristic index constructed by the combination of S-transformation and multi-fractal method can better identify the voltage sag source and effectively improve the classification effect of noise-containing signals and it can be applied to actual engineering.