Research And Implementation Of Harmonic Detection Algorithm Based On Convolution Window And Improved FFT

A large number of applications of nonlinear power electronic devices have caused harmonics to become the main factor affecting power quality.Accurate detection of harmonic parameters is the basis of harmonic analysis.The harmonic detection algorithm based on Fast Fourier Transform has good hardware portability and is widely used.This article focuses on the problems related to improving the accuracy of harmonic detection in the process of harmonic detection in power systems.The specific research content is as follows.(1)Based on the analysis of problems such as spectrum leakage and fence effect caused by non-synchronous sampling,the influence of sampling frequency and sampling length on the analysis results in windowing transformation is discussed,and the coordination relationship between the two values ??is clarified.The windowed and interpolated FFT transform can effectively suppress spectrum leakage.The sampling frequency and sampling length have a great influence on the analysis results.The reasonable range of the two values ??during signal detection is derived through formulas,and the influencing factors under different values ??are analyzed in detail.The four-spectrum interpolation algorithm is the object,and the analysis results are verified for the given complex harmonic signal.(2)Combining the discussion results of sampling frequency and sampling length,a cosine convolution window with better sidelobe performance is constructed,which improves the accuracy of harmonic detection.The relationship between the peak level of the sidelobe of the window function and the relative error accuracy of the amplitude shows that a window function with low peak level of the sidelobe and fast sidelobe attenuation can obtain higher harmonic detection accuracy.On this basis,a cosine convolution window with better sidelobe performance is constructed according to the convolution theorem.The influence of sampling frequency and sampling length in the convolution window on the spectral characteristics is studied.The experimental results show that the analysis result of the sampling length and sampling frequency are different can be extended to the convolution window.Finally,the correction formula of the bispectral interpolation algorithm with convolution window is given,and the signal detection accuracy of the proposed algorithm is verified through simulation experiments.The experimental results show that the overall signal detection accuracy of the bispectral line interpolation algorithm with convolution window is about 2 to 3 orders of magnitude higher than that of the unconvolution window,and the algorithm is still effective when the fundamental frequency fluctuates,white noise and inter-harmonic signals are detected.Better robustness.(3)Combined with the FFT algorithm,a software and hardware system for harmonic detection is constructed.This paper designs a harmonic detection system with TMS320F28335 as the main control chip,completes the design of related hardware circuits and software programs,and combines the upper computer to perform data display.Experimental results show that the system can complete the function of harmonic detection.