Extracting PD Signal by Applying the FPGA and Wavelet Transform
Xin LI,Wei-xuan HU,Ming-xia ZHANG,Cong-li ZHANG
DOI: https://doi.org/10.3969/j.issn.1003-6520.2007.08.002
2007-01-01
Abstract:Wavelet transform is one of the most effective methods for extracting partial discharge (PD) pulses from strong noise, however, since the complexity of computation, it is hardly used in PD online monitoring system and other high real-time fields when it is carried out on traditional embedded serial model microprocessors. While the FPGA (Fielded Programmable Gate Array) devices can work in parallel model and serial model, and it can be programmed to be special purpose chip, so it is more efficient in real-time field. In this paper, the Mallat algorithm of wavelet transform and wavelet package was introduced firstly, then, a scheme combining eclectic wavelet package with hard threshold processing was presented after analyzing the characteristics of PD signal and white noise at online transformer monitoring. According to this scheme, a pipe-line modularized wavelet transform system was designed subsequently on the basis of the FPGA using DSP Builder as the developing platform. An experiment of extracting imitated PD was done on this system, and the results prove that the system is feasible. At last, an experiment of drawing a real PD signal from a 220KV transformer was done, the results indicate that most of the real PD pulses are separated successfully, which is correspondent with the practice. It can be concluded that the system is feasible in field of extracting PD signal and other relative fields.