利用S变换与变尺度模板标准化的短时电能质量扰动分类

提出一种利用S变换与变尺度模板标准化的短时电能质量扰动分类方法。首先利用S变换对扰动信号进行时频分析;然后以最小二乘法选取最佳缩放尺度,通过变尺度方法标准化不同持续时间、幅值的扰动特征,在此基础上建立唯一的、时间尺度和频宽分辨率统一的各类扰动标准模板,作为识别扰动类型的直接判据;最后通过模板匹配的方法实现扰动分类。该方法不使用任何人工智能分类器,分类原理简单,过程明确,且可用于不同时间长度的扰动分类。Matlab仿真和实际试验结果显示,该方法能准确地对扰动进行分类且对噪声不敏感,是一种有效的短时电能质量扰动分类方法。

Classification of Short Duration Power Quality Disturbance Us-ing S-Transform and Scale Zooming Stan-dardization

A classification method for short duration power quality disturbance by adopting S-transform and scale zooming standardi-zation was proposed. First, power quality disturbance signals were decomposed with S-transform time-frequency analysis. Then, the best scale was selected by using least square method. All types of the disturbance charac-teristics with different du-ration as well as amplitude were standardized by scale zooming. Finally, the only standard templates with unified time scale and bandwidth resolution were constructed which could be used to directly identify the type of disturbance by template matching. The proposed method is able to achieve the different length of disturbance clas-sification without any arti-ficial intelligence classi-fier, so it is simple and clear. The results of Mat-lab simulation and actual examples show that the method has a thick skin to noise and is effective to classify disturbance sig-nals.