Research On Key Techniques For Distributed Online Monitoring Of Dielectric Loss Angle Of Capacitive Equipment

The measuring veracity and the measuring stability of dielectric loss angle of capacitive equipments are the precondition and basis to achieving from regular overhauling to condition maintenance. The thesis is financial supported by the key project'the online monitoring and diagnosing technology of the high voltage insulation device'(11511340)of province education department and the guiding project of province technology department.The thesis summarizes the highly accuracy measurement of tanδ. Based on the analysis of the correlation function method, least square method and the window harmonic analysis method, it shares the method of harmonic analysis method based on the wiener filtering of wavelet transform region. Through the extraction of the fundamental harmonic of the bus voltage and the leakage current, we can measure tanδ. Via the adaptive neural network and the wavelet transform to make the digital low-pass filter, we enhance the resolution of the leakage current. Through the simulation, compared with the normal harmonic analytical algorithm, the improved algorithm can restrain the leakage problem of asynchronous sampling caused by the fluctuation of system, and has higher accuracy.The thesis summarizes the method of data compression. Based on the analysis of the Huffman coding, the arithmetic coding and the adaptive Huffman coding, due to large amount and redundancy of the on-line monitoring data, we improve the embedded zero tree algorithm to compress the on-line monitoring data. Through the rational selection of the wavelet threshold, the monitoring information can be centralized effectively and the efficiency of the information management and further diagnosis can be improved.The thesis summarizes the extraction method of on-line monitoring signals. Based on the analysis of the mathematical statistics, the adaptive generalized morphology filtering and the wavelet singularity detection, we get the trend abstraction by the filter thought. In the frequency domain of the extracting monitoring signals, by the selection of the different threshold, wavelet decomposition and reconstruction, we reject the noise of the data. In the time domain, by the optimization correlative filter, we inhibit the additive noise and the modulation noise of the data.The grey relational analysis and the rough set theory are introduced to diagnose the insulation condition of capacitive equipments. We use the grey correlation degree to quantize of the similar degree among the different monitoring series and we can judge the insulation condition. By the synthesize of the support degree and the trust degree, we can excavate the general law of the insulation characteristic.The thesis summarizes utilizing the vector multiplier consist of the signal generator and the four quadrant multiplier to achieve the orthogonal decomposition of the leakage current signal. And we can achieve the filtering of the quadrant signals of the bus Voltage of leakage current, resistive leakage current capacitive leakage current to enhance the resolution of the leakage current. Combined with the synchronous sampling by the DSP, the influence problem of the dielectric loss precision caused by the power frequency fluctuation is reduced. In software design, to enhance the accuracy of tanδwe use the phase compensation of the spectrum leakage and the filtering phase shift, the on-line insulation monitoring and information searching system is achieved based on the Web.