Research On Wavelet Entropy Method For Power System Harmonic Detection

Along with the development of electric power system, the wide application of various nonlinear power electronics components and devices has brought about the increasingly serious power quality problem. As one of the important aspects of electric power quality problem, the harm that is caused by harmonic to the stable operation of the power system cannot be ignored. The power system harmonic detection, which is an important harmonic problem, is the research premise of the harmonic compensation and restrain. There are so many methods to detect the Harmonic. However, with the complication of the power system, it is very difficult to measure the power system harmonic accurately by using the traditional methods of the harmonic detection. Therefore, researching the new methods of harmonic detection is a more effective solution to the problem of harmonic.The dissertation first introduces the definition, source and harms of the power system harmonic and explains the importance of harmonic detection.Then through analysing the harmonic detection methods and implementation technologies both at home and abroad and comparing the differences among kinds of measuring methods and implementation technologies, we propose classified detection to the power system harmonic and transient disturbance signal by using the wavelet entropy method. After researching the theories of wavelet transform and entropy theory, the lifting wavelet energy entropy was constructed, and then used to detect the power system harmonic and transient signal. In order to prove the method, we use the Matlab to simulate the detection of the power system harmonic and transient signal by the wavelet entropy and the result is proved to be right.Second, In allusion to the disadvantage that the wavelet entropy method for harmonic detection fails to detect the steady-state harmonic's specific frequency and amplitude, we put forward the FFT operation of the local signal based the result of the wavelet entropy detection to improve the wavelet entropy method. On the basis of brief introduction of the Discrete Fourier Transform, the points'selection based the frequency of base-2 Fast Fourier Transform is importantly analysed. According to the wavelet entropy algorithm and FFT algorithm's complexity, we use the FPGA design the 64 points FFT processor, and through the software ModelSim and Matlab the FFT processor is jointly simulated. The results prove that the FFT processor design is feasible.Finally, wavelet entropy algorithm and FFT algorithm are implemented respectively by using the hardware of DSP2812 and Cyclone III EP3C25. And the experimental results ulteriorly verify that the feasibility of the wavelet entropy harmonic detection method.