Experimental Study On The Effect Of Roughness On Inversion Of SiO₂ Content In Iron Ore By The Thermal Infrared Spectrum

The accurate determination of the mineral chemical composition is of great significance for the development and utilization of mineral resources.Studies have shown that the Si-O bond has a strong fundamental frequency vibration in the thermal infrared band,and there is a good correspondence between the SiO₂ content in the rock and its thermal infrared spectrum.The use of thermal infrared spectroscopy to retrieve the SiO₂content in iron ore makes up for the shortcomings of traditional determination methods such as long time-consuming.In the evaluation of iron ore quality index and the formulation of smelting process,the SiO₂ content is a very important reference index.The determination of its content directly determines the smelting efficiency,heat loss and economic benefits.However,the thermal infrared spectrum of iron ore is affected by factors such as surface roughness,which leads to a decrease in the accuracy of SiO₂content inversion.Existing studies do not consider the influence of ore surface roughness on composition inversion,and use thermal infrared spectrum data to quantitatively invert the SiO₂ content in iron ore.The accuracy of inversion is difficult to provide for accurate delineation of the ore body range and ore matching effective help.Therefore,in this paper,the roughness is taken as a consideration factor affecting the inversion of SiO₂ content in iron ore,and the influence on the accuracy of inversion SiO₂ is studied.This article mainly researches the following aspects:1.Extraction of roughness of iron ore Fracture Surface.The representative"Anshan-style"iron ore was selected as the research object,and the shape of the fracture surface of the iron ore was obtained by scanning with a three-dimensional scanner.Use bilateral filtering and other algorithms to remove point cloud noise.Then the wavelet decomposition combined with the normal and fractal evaluation methods is used to obtain the surface roughness of the iron ore,and the distribution range of the surface roughness of the iron ore is calculated.The experimental results show that the roughness of the fracture surface of the 14 iron ore obtained basically conforms to the laws of normal distribution and fractal distribution.The roughness range is 2.29<Rq<9.74μm.2.Study on the influence of roughness on Inversion of SiO₂ content in iron ore.Based on the obtained roughness range of the iron ore fracture surface,and combined with the Becker criterion,two grades of roughness of 14 iron ore samples were produced.The surface roughness and thermal infrared spectrum of the sample were obtained by using a contact roughness meter and a thermal infrared spectroradiometer.The relationship between roughness and spectral characteristics and the influence of roughness on spectral characteristics were analyzed.At the same time,the roughness is taken as a factor that affects the accuracy of the inversion of SiO₂ content.The characteristic band that is most sensitive to the influence of spectral characteristics is selected and the normalization index is established.An inversion model of emissivity spectrum and SiO₂ content was further established to reveal the influence of roughness on the inversion results.The results show that:（1）The increase of the roughness Rq has a significant effect on the increase of the spectral emissivity of the RF（Reststrahlen Features）characteristic region.The average roughness Rq increased from 1.05μm to 2.47μm,which caused the maximum emissivity of the rough surface and smooth surface of the sample to increase by 0.17,a relative increase of 42.9%.（2）When the content inversion of the same grade of roughness is performed,the inversion error is small,the average relative error is 1.88%,and the inversion accuracy basically meets the error requirements of the geological and mineral industry standards.