Research On Harmonic Analysis Method Based On Wavelet Transform

The rapid development of power system has promoted the wide application of power electronic equipment and various non-linear loads.However,the problems of harmonic pollution and power quality caused by power system need to be properly treated.Accurate harmonic detection and analysis is an important basis for solving harmonic problems.The problem of harmonic detection in power system is that it is difficult to detect and analyze transient,abrupt and unstable signals.Real-time analysis and calculation of transient harmonic signals are the primary basis for evaluating and improving power quality.Aiming at the problem of poor robustness and accuracy of transient harmonic analysis in existing harmonic analysis methods,and this paper studies the merits and demerits of wavelet transform in harmonic signal processing,putting forward an improved harmonic analysis method of power system based on wavelet transform,solving the problem of time information lost by Fourier transform,and having the ability to detect sudden change,short-term impact,high-frequency resonance and other harmonic components with good effect.The main contents of this paper are as follows:(1)Analyzing the harmonic detection and existing problems in power system,focusing on the research status of harmonics in power system at home and abroad and the traditional harmonic detection and analysis methods,comparing the common methods of harmonic analysis at present,we find that the wavelet analysis method has unique advantages in signal analysis and processing,so we focus on the basic principles of wavelet transform and the wavelet theory of continuous wavelet transform,discrete wavelet transform,Mallat algorithm and wavelet packet algorithm,and make a theoretical analysis of the frequency aliasing phenomenon which inevitably occurs in the process of wavelet transform..(2)Comparing the advantages and disadvantages of the existing waveletanti-aliasing algorithms(aliasing compensation algorithm,interpolation optimization method),fully exploiting the characteristics of db series wavelet with large vanishing moment order and good frequency band division effect,and putting forward two-channel and two-stage aliasing compensation scheme to improve wavelet aliasing,the improved harmonic analysis method of power system based on wavelet transform can decompose signals into discrete wavelet transform,the decomposed sub-band signals are analyzed by continuous wavelet transform,and the wavelet coefficients are extracted,the sudden change frequency of the signal and the specific position of the signal in the time domain are obtained and the singular components are reconstructed,the simulation results show that the improved algorithm has good real-time performance and high accuracy.(3)The characteristics of harmonic signals in power system and the principle of singularity detection are analyzed;the anti-aliasing wavelet transform is used to analyze the common and transient harmonic signals in power system and the complex Morlet wavelet is used to analyze the time-frequency of time-varying harmonic signals,the wavelet transform is used to decompose the harmonic signals in power system and find the scale,the maximum modulus point of the local wavelet coefficients with the largest upper amplitude can accurately locate the starting and ending time of the sudden change point and obtain the amplitude and frequency information of the sudden change point,the simulations prove the validity of the wavelet transform in detecting the harmonic transient and the sudden change point of power system.In view of the abnormal situation of modern power system,such as abrupt change,short-term impact,high-frequency resonance and other harmonic components,FFT can not locate the occurrence time of harmonic signals because of its loss of time information;wavelet transform has variable time-frequency window,which is suitable for analyzing unstable signals,by improving the algorithm of wavelet transform and applying scenarios,it shows that the improved algorithm can effectively detect and segment unstable signals,the analysis of power system harmonics has certain practical value.