Extraction Of Partial Discharge Signal Feature With Chaos Oscillator And Adaptive Lifting Wavelet

The transformer default is majority caused by the influence of outside factor in operating, which makes the bug of transformer extending, and producing partial discharge(PD). The development of continues and acutely partial discharge leads to partial insulation aging of the transformer. So monitoring the PD is one of the mainly methods on on-line insulation monitoring of the transformer. However, those present PD monitoring the technologies are unsatisfactory for engineering application, it is because that pulse current is the common feature of all diagnosing methods, and noise restraining is the key trouble which haven't been resolved yet in online monitoring and diagnosing. Because of their excellent local time-frequency analysis characteristic and sensitive characteristic, adaptive lifting wavelet and chaos oscillator are widely used in signal analysis and feature extraction. Some useful and important research fruits of using them in PD online monitoring and diagnosing are obtained although too much key technology is still vacant.This paper studies the principles and methods of chaos oscillator and adaptive lifting wavelet in extending PD signal features, and obtains some valuable results.(1) The model of narrow-band interference signals detection based on Lyapunov exponents calculation and intermittent chaos is given.The chaotic motion of the Duffing oscillator is analyzed. It is concluded that the oscillator phase transformer is sensitive to the narrow-band interference signals which have the tiny angular frequency difference from the referential signal, and immune against the random noise and the partial discharge signals. The different frequency signal detection needs different frequency of referential signal, different frequency signal detection of the oscillator is universal by altering system parameters. The model of narrow-band interference signals detection based on Lyapunov exponents calculation and intermittent chaos is given. The applicability of the methods is verified by simulation analysis.(2) The lifting scheme of adaptive wavelet is improved.The lifting wavelet transform is employed to construct wavelet to solve the problem of white-noise reduction in PD detection. According to the theory and characteristics of the second-generation wavelet transform, the improved algorithm is proposed by using adaptive algorithm. The problem of nonlinear caused by adaptive algorithm is solved by using up-data first lifting scheme. To verify the efficiency of the improved lifting scheme, it is applied in the PD white-noise reduction and the results of de-noising by using the improved lifting scheme are compared with results of the traditional wavelet transform. The simulated results show that our improved lifting scheme has achieved the white-noise reduction better than that achieved by traditional wavelet transform.(3) A new method of de-noising is given.Because Duffing chaos oscillator immune against the random noise and the partial discharge signals, so first detect narrow-band interference in PD signals with chaotic oscillator, then de-noising with adaptive lifting wavelet. The applicability of the methods is verified by simulation analysis.