基于时频平面脊信息提取的电能质量扰动检测

Power Quality Disturbances Detection Based on Ridges of Time-frequency Plain

针对目前电能质量扰动检测方法的局限性,提出了一种基于时频平面脊信息提取的电能质量扰动检测识别新方法。该方法首先基于自适应最优核时频分布理论得到电能质量扰动信号的时频分布,然后提取了时频平面脊信息和基频脊幅值;结合时频平面的脊信息和基频幅值有效实现了对各种电能质量扰动的检测与识别。仿真测试结果表明:该方法表现出很高的时频聚集性,且不受交叉项的影响,清晰地刻画出各种扰动的时频分布特征,快速、准确地实现了各种电能质量扰动的检测与识别。

A new method of power quality disturbances detection and identification based on ridges of adaptive optimal kernel time-frequency distribution （AOK TFR） plain is presented. AOK TFR is a time-frequency analysis that can analyze unstable signal by nonlinear transforming in modern signal processing and the time-frequency distribution characteristics of all time varied signals can be revealed distinctly. Firstly, time-frequency distribution of power quality disturbances is acquired based on AOK TFR. Then ridges and fundamental frequency amplification of ridges are detected from time-frequency plain. Here practical fundamental frequency can be detected by frequency search, so the fundamental frequency amplification can be detected accurately. Therefore, all kinds Of power quality disturbances can be detected and identified. Some kinds of power quality disturbances such as voltage amplification disturbances, transient interruption, transient oscillation, time-varied interharmonics and SVC switching are simulated based on the method above. The testing results show that the method has higher time-frequency resolution, is independent from cross components, reveal time-frequency plain characteristics of all disturbances distinctly and the varied power quality disturbances are detected and identified accurately and fast. There is a fast algorithm for AOK TFR, and the core of the fast algorithm is based on FFT, detections of ridges and fundamental frequency amplify of ridges are so easy that their calculation burden can be neglected. Therefore, the method is suitable for practical application of power quality disturbance detection.