A Detection Method Of Voltage Fluctuation And Flicker Based On Frequency-searching

With the increasing power of fluctuating loads and the increasing useof frequency converter, the power quality issue caused by voltagefluctuations and flicker becomes more and more important. In order toreduce the harm to the end user, voltage fluctuations and flicker should beaccessed effectively. If the voltage fluctuation and flicker levels exceed areasonal value, they need to be effectively compensated. The detection ofvoltage fluctuation and flicker is an important precondition and basis forits compensation. Therefore, the study of effective detection method ofvoltage fluctuation and flicker has a very important significance.The square detection method recommended by IEC is a relativelymature voltage fluctuation and flicker detection method. This method iswidely accepted for it’s based on the light experiment. It can detect thevoltage fluctuation and flicker of a single frequency sine wave modulationsignal effectively, but for non-sinusoidal modulation signals, itsperformance is relatievely poor. In addition, similar to most traditionaldetection methods, IEC method is based on the amplitude modulationwave model which makes it difficult to detect the flicker caused byinterharmonics, especially for high-frequency interharmonics.To overcome the defects of IEC method, flicker caused byinterharmonics is analyzed in depth. Mehtods based on decompositionanalysis for the distort signal and beat frequency are discussed in detail.The peak fluctuation and RMS fluctuation of interharmonics are analyzedand the flicker curve of interharmonics is developed so that voltageflcutation and flicker can be calculated by interharmonics parameter directly. In addition, the relationship between amplitude modulation andinterharmonics is found so that a method that can apply for both amlpitudemudulation and interharmonics is found.In order to detect interharmonic parameters effectively, a methodbased on quasi-synchronization is proposed to improve the detectionaccuracy. Quasi-synchronization algorithm is realized by time-domaininterpolation of the sample sequence, and then using the comb filter toseparate harmonics and interharmonics, thereby reducing interferencebetween them. Then, a window is used to suppress the spectral leakage,and a bimodal interpolation algorithm in the frequency domain is used toget interharmonic parameters. This algorithm can detect the interharmonicparameters accurately, but the algorithm is limited by the frequencyresolution of the main lobe width of the window function, to improve theresolution of the algorithm, the length of the window function should beincreased, so that the amount of data increases and the algorithm efficiencydecreases at the same time.To solve this problem, a flicker detection algorithm based oninterharmonic frequency-searching is proposed based on thequasi-synchronization algorithm. The algorithm makes full use of thefrequency spectrum of the window function and finds the most likelyfrequency by assessing the likelihood between frequency spectrum ofwindowed interharmonics and the frequency spectrum of supposedinterharmonic frequencies. Then the result is calculated by a searchingprocess that aims to find the most likely frequencies. This method has avery good frequency resolution without changing the length of thewindow.Finally, IEC method, quasi-synchronous flicker detection method andthe frequency-searching based flicker detection method are simulatedusing Matlab respectively. The detection performance under differentcircumstances is evaluated, analyzed and compared in detail.