Research On The Rapid Detection Methods Of Voltage Sag

With the construction of smart grid, a large number of distributed generationshave been connected to the grid. It has made the characteristics and topologies of thepower system become more complex, and brought a large amount of power qualityproblems that cannot be ignored. In addition, the changes in the structure of the gridhave made large numbers of large-sensitive electrical and electronic equipment bewidely used in the power grid. Both the supply and customer put forward higherrequirements on power quality. Power quality problems have caused close attention ofscholars both at home and abroad. Because of its great harmfulness and highfrequency of occurrence, voltage sag has become one of the most serious transientpower quality problems. In order to guarantee the power quality, real-time monitoringand compensation of voltage sag is necessary. Dynamic voltage restorer(DVR) is oneof the most economical and effective methods for voltage sag compensation. Anddetecting voltage sag fast and accurately is the key and premise to make use of thecompensation device such as DVR to compensate voltage sag. This paper carried outresearch on the rapid detection methods of voltage sag in depth.Firstly, this paper briefly introduces the concept, causes, harms and suppressionmethods of voltage sag and so on. It focuses on analyzing with the voltage sagdetection methods. On this basis, through analyzing and comparing several kinds ofvoltage sag detection algorithms, we sum up the advantages and disadvantages ofvarious algorithms and their applicable scope.This article uses the good time-frequency localization features of wavelettransform, combines Mallat fast algorithm with wavelet transform modulus maximaprinciple to decompose a signal with multiresolution. And according to the principleof signal singularity detection, using the first and second scale of the high-frequencycoefficients of voltage sag after wavelet transform to locate the disturbances. Inaddition, according to the actual situation of voltage sag, this paper puts forward animproved soft threshold function for signal denoising. Simulation results show thatthe detection method of voltage sag based on wavelet transform can effectively detectdisturbances, but there are also some shortcomings, such as complicated algorithm, alarge amount of calculation, and the disturbances which take place in the moment nearn and the decline of their amplitude is not obvious that cannot be accurately positioned.Aiming at the shortcomings of the wavelet transform detection algorithm, thepaper presents a novel detection method of voltage sag based on Novel Teager EnergyOperator (NTEO) and RMS algorithm for voltage sag detection and measurementaccurately. In the proposed method, a resolution parameter i has been introduced toimprove the performance of TEO, then obtained NTEO, which has better anti-noiseperformance. First, the NTEO algorithm is used to locate the starting and ending timeof voltage dip disturbances, and then using the RMS algorithm to measure theamplitude of the positioned dips. The simulation experiment results indicate that, theproposed method has a stronger real-time property than wavelet transform, and itovercomes the lack of the poor anti-noise performance of TEO method and theshortage of RMS that cannot accurately locate the disturbances. The algorithm candetect the voltage sag fast and accurately, and fits the needs of DVR for real-timedetection.