Study On The Laser Plasma Spectral Characteristics And Measurement Methods Of Coal Particle Flow

Laser-induced breakdown spectroscopy (LIBS) is a potential analytical spectroscopytechnique for the quantitative analysis of the material composition based on the atomicemission spectra analysis. The significant progress has been made towards many industryfields due to its versatility of fast, in-situ and multi-element capability. Actually, the industrialprocess involves a lot of material in the form of particle flow. Therefore, it is very importantto apply LIBS to analyze the solid particle flow directly. The acquired information contributesto an important guiding for the development of industry detection system so as to realizeon-line detection of materials by LIBS. The determination of coal chemical composition priorto combustion is vitally important for a coal fired power plant to obtain optimal boilerperformance. The purpose of this thesis is to address the unique application of LIBS as a coalparticle flow monitoring technology, focused on the characteristics of coal particle flow andthe excitation mechanism between laser and coal particle flow in order to provide the basis forthe LIBS application in thermal power plants.The significance and background of this thesis were stated. The researches about LIBSapplied for the coal quality analysis and the composition measurement of gas solid flow werereviewed. Then the research contents of the thesis were proposed. In order to perform theresearch of this thesis, the basic experimental set-up and the cold state of simulation plateformwere established to realize the detection of the coal particle flow. And then the keyperformance parameters of the main components for the setup were detailed and the basiccharacteristics of laser-induced the coal particle flow were analyzed.According to analysis of the acquired spectra, it was found that the spectra containrepresentative spectra and partial-breakdown spectra. Spectral identification schemes wereused to acquire spectra containing all the emission lines of the important elements in coal,which were classified as representative spectra. The background of the line emission plusthree times the standard deviation of the background from the representative spectra waschosen as the threshold value. A single line method (C247.8nm，H656.28nm, N746.83nm)and a method with combined multiple lines (C247.8nm, N746.8nm, Si288.2nm, Ca396.8nm) were compared to obtain the best results for spectral identification of coal particle flow.The results showed that the single line method would keep all of the representative spectra butcannot reject all the partial-breakdown spectra, whereas the multiple emission lines combinedtogether could reject most (>98%) of all partial-breakdown spectra, which was found toprovide robust results for spectral identification of coal particle flow. The different pretreatment methods were compared to show that the RSD would decrease significantly byusing the normalized intensity and relative intensity and it can also reduce the number ofaveraged laser pulses.The laser ablation process has great relationship with the physical and chemicalproperties, of samples environmental conditions and the parameters of the laser. The deepstudies on influence of the above parameters on the laser-induced coal particle flow wereperformed with the actual demand of the measurement of the coal particle flow. The resultsshowed that the size of the pulverized coal flow beam, laser energy and wavelength would notonly affect the stability of plasma excitation but also cause the obvious influence of signalstrength. While for the influence from the plasma signal collection system, the main reasons isthat the nonuniform of the spatial distribution in the plasma and the limit of the signalcollection angle. On this basis, through the reasonable optimization of the above keyparameters, it could improve the accuracy of LIBS for direct measurement of coal particleflow significantly.Based on the understanding of the laser induced-induced coal particle flow plasma andthe coal particle flow characteristics, the related influence factors for the coal particledetection were summarized. The multiple linear regression (MLR) were used to establish thequantitative calibration model for the coal proximate analysis (calorific value, volatile matter,ash, fixed carbon) combined with the characteristics and the principle of the coal proximateanalysis. The results showed the root mean square prediction error (RMSEP) of thecalibration model method for calorific value, ash content, volatile matter, fixed carbon were1.73MJ/kg,1.73%,1.91%,2.4%respectively. The complicated chemical composition andphysical structure of coal and the characteristics of coal particle flow would be the mainreason to cause the fluctuation in the prediction for the measurement of coal particle flow byusing LIBS.