Study On Laser-Induced Non-metallic Elements In Solid Fuels And Quantitative Analysis Method

The combustion characteristics of solid fuels are closely related to the combustion process.The rapid and real-time detection of non-metallic elements and related industrial analytical indicators of solid fuels is great significance for the efficient and clean utilization of solid fuels.Laser-induced breakdown spectroscopy?LIBS?is gradually developing into a highly competitive industrial process for on-line detection of material components due to the advantages with no or simple sample pretreatment and multi-element synchronous rapid detection.However,the physical and chemical properties of the non-metallic elements and the various forms in solid fuels make it difficult to detect the non-metallic elements with LIBS.In this paper,laser-induced breakdown spectroscopy was proposed for the detection of non-metallic elements in solid fuels,and carrying out research work on the key issues involved in the plasma properties of non-metallic elements in solid fuels and the quantitative detection of the related elements and indicators.Firstly,the typical solid fuel is used as the research object.Under the two atmospheres of air and argon,combined with orthogonal experimental design,the single-factor effects and interactions of the critical parameters of the LIBS system on the relative standard deviation?RSD?,signal-to-noise ratio?SNR?and spectra intensity of non-metallic element?H,C,O,N?in solid fuels were investigated.The orthogonal table of L₅₀(5¹¹)was used for the experimental design.It can not only study the significance of the single factor effect of each parameter,but also analyze the interactions of delay time and laser energy,interactions of delay time and spot size on the LIBS spectra of non-metallic elements under the premise of fewer experiments.We builted an experimental platform for studying the temporal and spatial characteristics of laser-induced plasma of solid fuels,and the temporal and spatial distribution characteristics of plasmas of non-metallic elements H and N with different chemical forms are studied.The effects of different chemical forms on the spectral properties of plasma were obtained by correlating the plasma emission spectra of different chemical forms of H and N.Through the understanding of the temporal and spatial evolution characteristics of the non-metallic elements C,H,O,N in different carbonization coals and different solid fuels,the formation of various non-metallic elements was revealed.Different spectral pretreatment methods and variables selection methods were applied to correct the plasma emission spectra of coal and eucalyptus tablets,and the influence of spectral processing on the quantitative analysis model of calorific value and nitrogen content was analyzed.Normalization not only improved the stability of non-metallic elements,metal elements and molecular spectra,but also to some extent improved the correlation and predictive accuracy and precision of quantitative model.The matrix effect correction and spectral lines interference correction have significantly improved correlation coefficient of the calibration set,correlation coefficient of validation set and predictive accuracy and precision.The spectral variables selection method of interval partial least square?iPLS?further improved the performance of the model,especially for the prediction error and the RSD of predicted values.The studies have shown that high-precision quantitative analysis of solid fuel can be achieved through the spectral correction and iPLS variables selection.Aiming at the characteristics of LIBS spectra of pulverized coal particles flow,the influences of different spectral screening schemes on spectral rejection and spectral stability were compared and analyzed.The results show that the spectral screening of H,O,N combination not only made the invalid spectra rejection rate reach 100%and the effective spectra rejection rate is 0,but also obtained better spectral stability.Although the invalid spectra rejection and normalization improved the spectral stability of the pulverized coal sample,the coal property quantitative analysis model is still inferior.Therefore,the matrix effect correction and spectral lines interference correction method were introduced to correct the spectra of pulverized coal particles flow,and the spectral correction significantly improved the predictive accuracy and precision of the model.The spectral variables selection method of synergy interval partial least square?siPLS?was applied to further improved the prediction ability of the model and simplified it.