Chaotic Detection Method On Transformer Oil-dissolved Gas Photoacoustic Signal And Analyzing The Influential Factors

Gas photoacoustic spectroscopy as a new sensing technology,it has not consumption or separation gas,high sensitive,good selective,large dynamic range etc excellent features. It provides new ideas and technologies for gas dissolved in transformer oil detection. Photoacoustic signal gas is the only characteristic for gas concentrations analysis, also is the key of the detection system.Firstly,this paper introduces the stimulate mechanisms of gas photoacoustic signal in detailed,emphatically describe the acoustic field distribution and photoacoustic signal amplitude of cylindrical photoacoustic cell, Based on the theory and method of photoacoustic spectroscopy,author designs the gas photoacoustic spectroscopy testing platform, detailed introduces the photoacoustic cell design and production, and analysis the noise of detection system.Aiming at the limitations of currently method to gas photoacoustic signal detection,this paper introduced a conventional chaotic measuring technology. Detailed analysis the detection principle of duffing chaotic system,establish the simulation model for frequency known sine periodic signals,and analysis of the system nonlinear dynamic behavior,Based on this,puts forward a photoacoustic signal of gas in oil detection method based on the cross-correlation and chaotic,and use two kinds of mathematical tools judge state,maximum lyapunov method and double chaotic oscillator difference method analysis on gas photoacoustic signal,the results error not exceed 4.43% and 5.95% respectively,which show that the proposed method can restrain noise detect the weak sine signal in strong noise,and two method have their respective advantages.Besides this paper analyzed the affect factors of gas photoacoustic signal and discuss the relationship between pressure,temperature,laser power,gas concentration, resonant frequency, background gas,resonant frequency drift and gas photoacoustic signal value in detailed, provides the technology reference and support for photoacoustic spectroscopy detection system.Considering the linear relation between single-component gas photoacoustic signal and gas concentration,a photoacoustic spectroscopy quantitative analytical method based on the least squares regression is given,and establish the quantitative analysis model.Use this method to inversion gas concentrations can improve the calculation accuracy.Experimental analysis results with CH4,C2H2 gas shows that the difference of this method and gas chromatographic is not big, and the error of CH4 gas detection does not exceed 4.75%,error of C2H2 gas detection does not exceed 4.22%.