Analysis And Improvement Of Modern Method And Test Platform Design For Harmonic Detection

With the continuous development of power electronic technology,nonlinear equipments are widely used.The increasingly serious harmonic pollution problem in distribution network poses a threat to the security,stability,high-quality and economic operation of the power system.Therefore,the accurate detection and analysis of harmonic become the premise of harmonic elimination.This paper carries on the comparative analysis of harmonic modern detection methods based on Fourier transform,Wavelet Transform and Hilbert-Huang Transform(HHT),and determines the advantages of the HHT detection method.On this basis,this paper analyzes the mode mixing problem of the empirical mode decomposition(EMD)in HHT,and adopts the ensemble empirical mode decomposition(EEMD).The simulation shows that this method overcomes the effect of intermittency signal on the EMD and has relatively improved the decomposition ability and the anti-interference ability.At the same time,aiming at the practical application of harmonic detection,the conclusion can be drawn that the low SNR signal and the dense frequency signal of the high harmonic affect the EEMD.Based on the above two kinds of problems,thorough research is carried on in this thesis.Firstly,the improved wavelet threshold de-noising method and the improved EEMD threshold de-noising method are put forward to completing the noise suppression preprocessing of the low SNR signal.It has been verified by tests that both kinds of de-noising method can achieve a very good de-noising effect.After comprehensive consideration and analysis,the improved wavelet threshold de-noising method is selected for the noise suppression preprocessing in this paper.Secondly,aiming at the mode mixing problem of the EEMD brought by the dense frequency signal of the high harmonic,the decomposition ability of the EEMD is tested and analyzed.The EEMD harmonic detection method based on the frequency shift of Hilbert is proposed.The test shows that this method can separate the high harmonic accurately,and effectively overcome the mode mixing problem brought by the dense frequency signal of the high harmonic.By combining de-noising method and the harmonic detection methods proposed in this paper,this paper improves the applicability of the EEMD in harmonic detection.Finally,this paper designs the harmonic detection test platform to provide an open platform for harmonic detection method.The harmonic detection system with the industrial computer as the core is the master computer of this platform.By hybrid programming with LabVIEW and MATLAB,this system is used to test the de-noising method and the harmonic detection method proposed in this paper.The effectiveness of the method proposed in this paper is further proved by the test.In addition,with full consideration of engineering practicability for detection method,a set of harmonic detection system with the DSP28335 chip as the core is served as the slave computer of this platform and the design of its hardware and software is completed.It is verified by the test that this system can realize the detection and analysis of harmonic.