Photoacoustic Spectrum Detecting And Insulation Diagnosing Technology Based On Transformer Oil Dissolved Gases Analysis

Large power transformer is one of the most important and expensive equipments in power industry. Its operation reliability involves the security of power supply and has significant impact on national economy development and people's lives. With the development of power system in the direction of extra-high voltage, large power grids, large capacity and automation, the research of large power transformer's insulation detecting and fault diagnosis technology is carried out to detect the latent failure of power transformer insulation. This plays great theoretical significance on the safe, reliable, stable operation for transformer and even the entire power system. There are two primary problems existing in the detecting devices of dissolved gases in transformer oil: one is that the precision and long-time stability of sensor is not high, and the other is that the chromatogram column doesn't fit continuously on-line monitoring. To solve these two problems, we need to choose new-style gas-detecting technology.This paper starts with application of photoacoustic spectrum in gas detecting. Aiming at this point that hydrogen has no absorption in the infrared band; this paper gives the qualitative analysis of relation character between the concentration of hydrogen and the phase of photoacoustic signal based on the principle that the concentration of hydrogen can influence the phase of photoacoustic signal. Based on the bringing mechanism of photoacoustic signal, the relation character between signal value and concentration of other characteristic gases is achieved.Considering the current research status of detecting technology of gas photoacoustic spectrum and the requirement of application on operation scene of transformer, this paper represents composing components of test platform used in detecting dissolved gases in oil and analyzes the principle and design realization of the pivotal component-photoacoustic cell. According to the strong noise interference in detection process of photoacoustic signal, this paper combines wavelet threshold denoising with chaotic detection to restrain strong noise. And use LabVIEW assorted with Simulink simulation model to set up virtual signal detection system to pick up weak photoacoustic signal. The experimental results represents this system is very sensitive to weak photoacoustic signal under strong noise, it can restrain strong noise in the detecting process effectively and improve signal-noise ratio in a great degree. Finally, the shortcomings of BP algorithm used in diagnosing techniques are analyzed and wavelet neural network is introduced. The weakness of wavelet neural network parameters optimized by BP algorithm (BPWNN) is analyzed and wavelet neural network parameters optimized by adaptive genetic algorithm (AGAWNN) is proposed to overcome the weakness. A comparison between BPWNN and AGAWNN is made by using the transformer oil dissolved gas data. The results show that AGAWNN has better performance than BPWNN in the speed of convergence, structural optimization and diagnostic results and so on.