Research On Detection Of Primary Air Coal Powder Based On LIBS

Coal is the most widely used primary energy source in thermal power generation in China.The optimization of coal proportion can effectively improve coal combustion efficiency.The analysis of coal quality characteristics is an important means to improve combustion efficiency and reduce coal consumption.The existing coal quality analysis technology mostly adopts the traditional off-line analysis method,which has long detection cycle and low efficiency,and is difficult to realize the role of real-time on-line guidance of furnace combustion.Laser induced breakdown（LIBS）technology has many advantages,such as simultaneous measurement of multiple elements,wide application range and fast analysis speed.It can be applied to on-line detection of pulverized coal composition.Aiming at the problems of LIBS commonly used pulverized coal flake detection,such as inadequate real-time performance and inability to truly reflect the characteristics of coal into the furnace,this thesis carries out on-line detection of pulverized coal particle flow.The research content is as follows.Firstly,because of a comprehensive understanding of the research status of LIBS technology at home and abroad,LIBS is applied to the detection of pulverized coal composition in primary air pipelines of power plants,and a detection platform for pulverized coal particle flow is built.The detection platform was mainly composed of a pulverized coal conveying system,a laser induced system and a light receiving detection system.The speed of pulverized coal flow in the experimental platform was controlled by PWM speed control module.Laser excites the pulverized coal flow.The excitation spectrum of pulverized coal was collected by spectrometer.These data wre used to analyze the composition and concentration of pulverized coal elements.Secondly,due to the matrix effect of pulverized coal particles,the peak attenuation of spectral lines will affect the detection accuracy.In this thesis,a method to compensate the fluctuation of the spectral intensity of the main elements in the pulverized coal particle flow was established.The plasma temperature and electron density were calculated to compensate the fluctuation of the physical characteristics in the pulverized coal particle flow.Using C₂ and CN molecular spectra to compensate the light intensity of C element detection,at the same time,the attenuation of peak value of main element spectra in coal was corrected,and the utilization rate of light intensity of main element in coal was improved.Finally,in the study of quantitative analysis of major elements in coal,the partial least squares（PLS）non-linear multi-spectral dominant factor method was introduced into the modeling and analysis of the determination of major elements in pulverized coal particle flow.This method can effectively solve the problem of excessive residual caused by the dominant factors such as self-absorption effect,mutual interference effect and fluctuation of plasma parameters.Compared with the traditional calibration method,the PLS fitting non-linear multi-variable dominant factor model has a good prediction ability for the detection of pulverized coal particle flow.The goodness of fit for the calibration and prediction curves of the main elements has been improved,the average relative error of sample calculation has been reduced,and the indicators have shown a good trend.The results show that the LIBS detection platform built in this thesis can simulate the detection function of pulverized coal concentration in primary air pipeline,eliminate the matrix effect of the main elements in coal,combine with the dominant factor of PLS non-linear multi-spectral line to realize the prediction function of the concentration of the main elements in pulverized coal flow,and provide a theoretical basis for the application of LIBS-based pulverized coal particle flow detection technology to coal quality detection in power plants.