Study On Spectral Characteristics Of Pulsed Discharge Plasma Of Pulverized Coal Prepared By Laser Pre-ablation

The fast online measurement of coal quality is beneficial to the effective and real-time guidance of the boiler in the power plant,which is very important for the economic benefit of the power plant.Laser Induced Breakdown Spectroscopy(LIBS)has been used for elemental analysis and industrial analysis of coal because of its advantages such as no or simple sample preparation,multi-element analysis and in-situ detection.However,when high-energy laser ablates coal samples in LIBS,it would lead to serious matrix effects caused by volatilization and combustion of volatiles in coal,which seriously limits the measurement accuracy.In this paper,a new measurement method,Laser Ablation Spark Induced Breakdown Spectroscopy(LA-SIBS),is proposed for coal quality analysis.LA-SIBS uses a low-energy laser to ablate the coal sample,and the sample produced by the ablation is excited by a high-voltage pulse to generate a plasma.In this way,the matrix effect of high energy laser ablation can be avoided.In this paper,a set of LA-SIBS device for coal quality analysis was built,and LA-SIBS technology was applied to coal quality analysis.In order to improve the sensitivity and repeatability of the detection,the effects of delay time between pulse and spectrometer,electrode material,delay time between laser and pulse,the lens position,laser frequency and laser energy on the spectrum were analyzed and the relevant parameters were optimized.The optimal experimental parameters in this experiment are as follow:the delay time between pulse and spectrometer is 20?s,the electrode material is tungsten,the delay time between laser and pulse is 100?s,the lens position is+1.5mm(+represents the anode direction of the electrode),the laser frequency is 3Hz,laser energy is 30mJ.Based on the above parameter optimization,the performance of LA-SIBS and LIBS for coal quality analysis was compared and analyzed.The partial least squares method is used to model the spectrum.By comparing the model's fitness,root mean square error of prediction,average absolute error of prediction and average relative error of prediction,the measurement effects of two spectral techniques on the carbon content,volatile and ash content in coal were compared.Under LA-SIBS conditions,the linear correlation coefficient R~2,average absolute error of prediction,the average relative error and the root mean square error are:carbon content:0.977,1.00%,1.53%and 1.24%;volatiles:0.998,1.43%,10.34%and 1.77%;ash:0.998,0.86%,3.17%and 1.37%;under LIBS conditions,the linear correlation coefficient R~2,average absolute error of prediction,average relative error of prediction and root mean square error of prediction are:carbon content:0.993,1.52%,2.49%and 1.68%;volatile matter:0.948,1.78%,12.87%and 2.16%;ash:0.999,1.32%,5.83%and 1.40%.Root mean square error of prediction,average absolute error of prediction and average relative error of prediction under LA-SIBS conditions are lower than LIBS conditions,so the prediction effect of LA-SIBS is better.