Research On Soft-sensing Model Of Premixed Hydrocarbon Flame Equivalence Ratio Based On Color Thicknes

Environmental pollution and the clean and efficient use of energy are hot issues of concern and scientific research in today’s society.Under the rapid economic development,the demand for energy is increasing day by day,which makes the environmental problems more and more serious.However,China’s energy structure makes China still need to rely on fossil fuel combustion in the long term,in the process of energy conversion of fossil fuels,combustion efficiency is an effective indicator to characterize the utilization of energy,and improve the combustion efficiency can promote the effective use of energy and reduce the emission of pollutants.The efficiency of gas combustion can be characterized to a certain extent by the equivalence ratio,and the detection and control of the equivalence ratio can effectively improve the utilization of fuel and improve the environment.It has been shown that the hydrocarbon flame equivalent ratio is closely related to the flame radicals during combustion,and the flame equivalent ratio can be obtained indirectly by measuring the flame radicals CH^（9）and C₂^（9）.With the high speed development of machine vision,it has been found that the ratio of B and G layers of flame image has similar change withCH^（9）/C₂^（9）.The color measurement method for the equivalent ratio was established on this basis,and a soft measurement model of the equivalent ratio based on flame color was constructed by studying the flame color-emission spectrum-radical-equivalent ratio relationship.In the traditional color method,flame color information needs to be obtained through an image sensor.According to the principle of two-dimensional image imaging,the two-dimensional projection image of the flame is a line integral of the flame luminescence over the three-dimensional spatial depth information,which makes the acquired two-dimensional flame color exist in thickness,i.e.,the influence of spatial depth information exists.Therefore,this paper proposes to investigate the effect of color thickness on the soft measurement model of premixed hydrocarbon flame equivalent ratio from two perspectives of removing the flame color thickness and recovering the distribution of color thickness in space,respectively.Firstly,methane flame and propane flame were used as the research objects,and the flame image data under different working conditions were collected,and the image data were filtered for the subsequent research needs.For the data of partially mixed premixed flame and diffusion flame,the digital flame color distribution technique was used to realize the separation of premixed flame and diffusion flame,and finally the data of flame region were extracted.Secondly,this paper adopts Abel’s inverse conversion method to eliminate the color thickness of the flame,and based on this comparison with the traditional method,it is concluded that the color thickness has some influence on the traditional soft measurement model,and the color thickness has more influence on the methane premixed flame than on the propane premixed flame.For this situation,six different Abel inversion solutions are used to further investigate methane premixed flames in this paper,and after comparison,it is concluded that the direct method is better than several other methods when applied to methane premixed flames.Finally,this paper uses three-dimensional reconstruction technology to reconstruct the color thickness of the flame in three dimensions,and on the basis of this study,the equivalence ratio changes at different depths at different times,and it is found that the equivalence ratio of the local area of the flame changes with the depth of the flame,and it is found that the equivalence ratio of the local area of the flame increases with the increase of the flame depth,which also proves that the color thickness will inevitably have an impact on the model of measuring the flame equivalence ratio based on the color method.