基于S变换图像灰度量化的短时电能质量扰动识别分类

针对短时电能质量扰动分类大多依赖分类器,分类准确率不高这一难题,提出了基于S变换模时频矩阵灰度图像法。首先对常见的几种扰动进行S变换分析,得到模时频矩阵,再应用数字图像灰度方法,将模矩阵各元素值用灰度图方式表示,分析其灰值分布特征,引入灰度期望和灰度方差两指标,量化灰度图像灰值分布,并根据量化结果建立扰动标准判据,实现扰动分类。仿真实验表明,该方法不依赖于分类器,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分类方法。

Identification and Classification of Short Duration Power Quality Disturbance Using S-transform and Grey Scale Method

A classification method for short duration power quality disturbance by measuring the grey scale image of S-modal time-frequency matrix is proposed.Firstly,the common power quality disturbance signals are decomposed with S-transform analysis and the S-modal time-frequency matrix is constructed,the grey scale image of which could be easily calculated according to digital image grey scale method.Then,the grey value and grey variance are introduced to quantify the characteristics of S-modal time-frequency image in order to achieve the disturbance classification.The simulation results show that the proposed method can classify the disturbances exactly and is not sensitive to noise,so it is an effective classification method of short duration power quality.