Fundamental Research On Real-time Coal-rock Interface Recognition Using Terahertz Technology

At present,the position of coal in the energy market cannot be completely replaced in the short term,and China's shallow coal resources are nearly exhausted.However,the characteristics of high osmotic pressure,high in-situ stress,and high ground temperature in deep formations have caused the incidence of coal mine disasters to increase sharply.In order to reduce casualties during coal mining,some scholars have put forward the concept of'unmanned coal mining'.The realization of unmanned fully mechanized mining face requires a lot of technical support.One of the key technologies is the online recognition technology of coal-rock interface.Terahertz technology has achieved rapid development in the past thirty years and has shown great application potential in many fields.The use of terahertz waves to identify the coal-rock interface is still in the early stage to be explored.The research of terahertz technology in substance identification and quantitative analysis is becoming more and more mature,but it has high requirements on the form of the sample to be tested,and the time-consuming and cumbersome sample preparation process limits its application in occasions with high requirements for real-time performance..On-line identification of coal-rock interface requires that coal and rock can be identified in real time,quickly and accurately.A series of basic research on the online recognition method of coal-rock interface was proposed based on terahertz radiation.The principle framework of using transmission THz radiation to identify the coal-rock interface online is proposed,focusing on: how to quickly distinguish and identify granular coal and rock,and how to identify the problem of coal particles mixed with a small amount of rock.Some basic research has been done.The main research work and innovations of the dissertation include:(1)Transmissive terahertz radiation was used to identify the principle model of coal-rock interface online,and it is expected to solve the problem of online coal-rock interface identification.This principle model collects the dust and falling objects generated during the coal cutting operation of the shearer,and obtains particles with a specified particle size range through crushing and separation,and then detects these particles with terahertz radiation to realize the classification and identification of coal and rock.This principle model saves more than ten minutes of sample preparation time and improves the real-time performance of coal and rock identification.(2)The transmittance model of coal and rock particles was established,and the influence of the particle size of coal and rock on its terahertz spectrum was explored experimentally.In response to the question: how the particle size of granular coal and rock affects its terahertz spectrum,the dissertation measured and extracted the optical parameters of 8 coal and rock samples in the terahertz frequency band to theoretically calculate the transmittance of granular coal and rock Provide data basis;establish the transmittance model of coal and rock particles based on Lorenz-Mie Theory;experimentally analyze the terahertz spectra of 8 kinds of coal and rocks in the particle size range of 1998?m-8?m,and observe the deformation of the terahertz spectrum caused by the particle size,Distortion,etc.The results show that: the overall trend of the transmittance measured by theoretical calculation and experiment is consistent;the transmittance of coal and rock particles is a function of particle size and refractive index;inferred that granular coal and rock can be distinguished according to the transmittance spectrum.(3)A two-dimensional parameter map method based on the transmittance of three frequency points was proposed,which realized the distinction between granular coal and rock in the 1998?m-8?m particle size range.In response to the problem: how to identify the distorted terahertz spectrum of coal and rock particles,the dissertation uses the parameters calculated by the transmittance of three frequencies to draw a two-dimensional parameter map,and realizes the distinction and identification of coal and rock particles.Moreover,the specific combination of particle size and frequency can also be used to distinguish the types of coal and rock.This method proves that only three frequency monochromatic terahertz laser sources and detectors can realize the identification of coal and rocks in the 1998?m-8?m particle size range.Compared with using a terahertz time-domain spectrometer for qualitative analysis of pressed mineral samples,this method of using a monochromatic terahertz laser source and detector to detect powder-form substances is faster and cheaper,and can better meet the real-time requirements of the industrial field.And lowcost demand.(4)Support vector machine was used to classify terahertz spectrum of granular coal and rock.In response to the problem: how to identify the terahertz spectrum of distorted coal and rock particles,the dissertation proposes a second method.The dissertation uses support vector machines to classify the terahertz time-domain spectra of coal and rock;explores which spectrum is used as the support vector machine data set,which can classify coal and rock particles with higher accuracy;and explores the classification accuracy of particle size Impact.The experimental results show that: choosing the time-domain spectroscopy data of particles with particle size levels 1-8(average particle size within 1142?m-212?m)as the SVM data set,better classification accuracy can be obtained,and the average classification accuracy rate can reach 86 %above.This provides a basis and basis for the formulation of sample particle size standards and the selection of data sets in the principle model of online identification of coal-rock interface.(5)The transmittance model of the binary particle mixture was established,and the experiment explored the influence of the mixing ratio on the transmittance of the coal-rock mixture.Being able to identify a small amount of rock mixed in coal is essential to improve the response speed of coal-rock interface recognition.In response to the question: How does the mixing ratio of the coal-rock mixture affect its spectrum? Taking the mixed powder of carbonaceous mudstone and anthracite as an example,the transmittance model of binary mixtures with different volume fractions is established based on Lorenz-Mie Theory;Obtain the transmittance of the mixed powder of different mass ratio'anthracite-carbonaceous mudstone'.The results show that the transmittance calculated by the theoretical model is very consistent with the experimentally measured transmittance,the values are close,and the trend is the same;it shows that the transmittance of the coal-rock mixed particles is a function of the mixing ratio;inferred that the coal-rock mixture can be quantitatively analyzed based on the transmittance spectrum.(6)Using the two-dimensional parameter map method,the quantitative analysis of coal-rock mixed powder is realized.In response to the problem: how many rocks can be mixed to make the coal-rock mixture identifiable.The dissertation takes carbonaceous mudstone and anthracite as examples,uses three transmittances at 200 GHz,500GHz,and 800 GHz to calculate and draw a two-dimensional parameter map,and realizes the calculation of coal and rock Quantitative analysis of the mixture;and observed that when the mass fraction of carbonaceous mudstone reaches 20%,its parameters can be distinguished from those of pure anthracite.The results show that the transmittance of only three frequency points can be used to quantitatively analyze the'carbonaceous mudstone-anthracite' mixed powder with a resolution of 15%-20%;when the carbonaceous mudstone mixed in the anthracite reaches 20% of the total mass %,the particulate matter can be determined as a mixture of'carbonaceous mudstone-anthracite' through the two-dimensional parameter map.The research results of the dissertation are helpful to the application of terahertz technology in the field of online recognition of coal-rock interface,and promote the qualitative and quantitative analysis of terahertz technology for substances into industrial application scenarios that have high requirements for real-time performance.