The Analysis Of Power Quality Disturbances Based On Time-frequency Methods

Global industrialization developed more and more rapidly,and non-linear equipment,such as VFD,they are widely used.It makes people's lives more convenience,at the same time,it also brings some power quality problems such as voltage interruption,voltage sags,harmonics and other power quality problems.The key to improve the power quality is that a good detection method is needed to provide accurate and detailed parameters of power disturbance.Therefore,it is very important to accurately detect the eigenvalue of the disturbance signal and realize the classification and recognition of the signal.This paper takes MATLAB software as the simulation platform and establishes seven signal models.It includes one ideal sinusoidal voltage signal,four single disturbing signals,such as harmonics,voltage temporary rise/sag,interruption.And two compound disturbance signals such as harmonics voltage rise / fall.Using time frequency analysis methods to simulate the above models.Firstly,the wavelet transform and short time Fourier transform are used to simulate the voltage sag that with and without noise and the harmonic signal.By comparing the simulation results of the two methods,it can be concluded that the wavelet has the error in the analysis of the sag signal with noise,and can not accurately extract the eigenvalue of the temporary disturbance.On the contrary,the short-time Fourier transform has better anti-jamming ability,and can detect the signal amplitude,frequency components,the starting and ending time of the signal clearly and accurately.Therefore,the short-time Fourier transform is superior to wavelet in the aspect of anti-noise performance and amplitude detection.Then,using the short-time Fourier transform to simulate the above seven disturbance models.This paper proposed the method that determination of the peak position of the envelope in the low frequency amplitude envelope in the positioning of the disturbance time,and detection of frequency component of signal by amplitude frequency curve,and the change of fundamental frequency amplitude of fundamental frequency amplitude curve.Finally,the simulation results of short time Fourier transform are analyzed in detail.In this paper,the classification method of decision tree is used to classify the disturbing signal.The principle of classification is based on the simulation results of the energy disturbance signals which transformed by the short-time Fourier transform and the characteristic quantities used to disturb the classification are extracted from the amplitude curve,the fundamental frequency amplitude curve and the amplitude time-frequency matrix respectively.Decision tree and the amplitude characteristics of each disturbance signal to classify it.,combined with the characteristics of the decision tree and the amplitude of the disturbance signals,the disturbance characteristics are transformed into the characteristic quantities which are related to the amplitude,Through the analysis and verification of multiple sets of data,this classification method can classify the disturbance signal,and the accuracy is simple,the principle is simple and easy to implement.