Spectroscopy Analysis And Detection Technologies Of C₄F₇N/CO₂ Gas Mixture And Its Decomposition Products

As an excellent insulating medium,SF₆ gas is widely used in power industry,but its strong greenhouse effect is always a threat to the global atmospheric environment.In order to further strengthen green development and actively respond to climate change,the research on environmentally-friendly gas insulating medium has attracted more and more attention.Under the general trend of global green development,the related technical problems in the development and application of environmentally friendly gas-insulated electrical equipment have become important research topics at the moment.C₄F₇N gas is currently the most environmentally friendly insulating gas with the greatest potential to be used in high-voltage electrical equipment.With the popularization and application of the C₄F₇N/CO₂ gas mixture insulation equipment,it will inevitably face operation and maintenance problems such as equipment condition assessment and fault diagnosis.This paper starts from the research of the C₄F₇N/CO₂ gas mixture detection technology and gives full play to the application advantages of spectroscopy detection technology in the power industry.Spectroscopy analysis of the C₄F₇N/CO₂ gas mixture and its decomposition products is carried out,and related detection schemes based on spectroscopy detection technology are proposed.（1）Based on time-dependent density functional theory,theoretical calculations are performed on the ultraviolet（UV）spectrum of the C₄F₇N molecule.It is found that the UV spectrum of the C₄F₇N is mainly concentrated in 100～200 nm.The UV spectroscopy experimental platform with a light source band of 185～200 nm has the feasibility to detect the C₄F₇N gas,which can effectively avoid the interference of CO₂ gas.Theoretical and experimental analysis of the UV spectrum characteristics of the C₄F₇N gas at 185～200 nm is carried out,the results show that the UV absorption peak in this band is caused by the electron transition from HOMO to LUMO,with the contribution of 91%.The interference effects of the UV spectroscopy experimental platform are analyzed.By selecting the area of the right half of the absorption peak as the characteristic value,the interference effects can be effectively eliminated.The ratio detection of the 4～10%C₄F₇N/CO₂ gas mixture with an inversion error of less than 5%can be achieved.The performance of the UV spectroscopy experimental platform is analyzed,its detection limit is 0.37%,and it has good detection stability.（2）The difference of approximation methods of different exchange–correlation energy functional in the optimization of the C₄F₇N molecular geometric structure is analyzed.Combined with the FTIR experimental platform,the IR spectrum characteristics of C₄F₇N have been analyzed,the results show that the hybrid functional B3LYP is more accurate for the calculation of the IR spectrum of C₄F₇N.Based on this method,the vibration mode of the main IR absorption peak of the C₄F₇N gas is analyzed.The IR spectrum of the C₄F₇N gas is distributed in 400～2400 cm^-1,among which the IR absorption of 900～1400 cm^-1 is strong,and a large area of IR spectrum saturation will appear under high mixing ratio.The IR spectrum distribution characteristics of the C₄F₇N gas under high mixing ratio are analyzed,and it is found that the spectral overlap peaks of 750～785 cm^-1 are suitable for mixing ratio detection.The Gaussian function fitting algorithm has been improved.Using multi-scale wavelet transform to obtain the peak number and approximate peak position of the IR spectrum overlapping peaks,which can provide a limited condition for the Gaussian function curve fitting algorithm and it will optimize the time complexity of the algorithm.The iterative adjustment idea of particle swarm optimization is used for reference,and the inertia factor is introduced in the solution process to ensure the stability of convergence.After splitting the overlapping peaks in the 750～785 cm^-1 band,the mixing ratio inversion error can be reduced from 4.8%to 2.7%.（3）The geometric structure and the IR spectrum of the decomposition products of C₄F₇N are theoretically calculated.The experimental values of the geometric structures of some decomposition products are compared with calculated values to verify the accuracy of the calculation method.The feasibility of IR spectroscopy detection of the decomposition products is theoretically analyzed.The C₄F₇N/CO₂ gas mixture overheating decomposition experimental research is carried out.The long path gas cell combined with IR spectroscopy is used to qualitative detection of decomposition products,the results show that this method can effectively detect the typical overheating faults decomposition products of C₃F₆,CO and COF₂.By analyzing the absorbance value of the subtractive spectrum,the generation trends of the three decomposition products of C₃F₆,CO and COF₂ at different temperatures and pressures are summarized.The concentration of CO gradually increases with the increase of overheating temperature.The formation of COF₂ occurs at 450°C,and its concentration does not increase significantly when the temperature reaches 500°C.The concentration of C₃F₆gradually rises with the increase of temperature when the overheating temperature is lower than 450°C.When the temperature reaches 500°C,the concentration of C₃F₆ drops sharply and then slowly rises.At the same temperature,as the pressure increases,the concentration of the three decomposition products will gradually decrease.（4）The quantitative analysis method of C₄F₇N and its decomposition products is carried out.100 sets of gas mixture samples containing decomposition products are randomly prepared,and the IR spectrum data of the samples are obtained using the long path gas cell with FTIR to construct the IR spectrum sample library.A combination of multiple spectral preprocessing methods and three typical chemometric algorithms is selected to establish a chemometric quantitative analysis model,and finally the best IR spectral preprocessing method and chemometric algorithm are determined.The results show that the best modeling method is PLS,the best spectral preprocessing method for C₃F₆ and C₄F₇N is no processing,and the best spectral preprocessing method for CO is Savitzky-Golay smoothing filtering.The error distribution of the quantitative method is analyzed,and the internal cross-validation proves the quantitative analysis ability of the model.The quantitative analysis results of decomposition products under different overheating conditions showed that the CO concentration did not increase significantly when the temperature reached 500℃.The generation rules of C₃F₆ and CO under different pressures are relatively close.When the air pressure rises from 0.15 MPa to 0.3 MPa,the generation amount is reduced by more than half,and then as the pressure continues to increase,the reduction amount decreases.This paper has perfected the research work of C₄F₇N gas from the characteristics of the UV spectrum and the IR spectrum.The detection schemes of the C₄F₇N gas and its decomposition products have been further proposed.It provides theoretical support and technical guidance for the application of spectral detection technology in the operation and maintenance of the C₄F₇N/CO₂ gas mixture electrical equipment.