Fast Detection Of Pulverized Coal Composition Based On LIBS

Coal as the proportion of our country’s first energy,its position is very important.Development of fast and on-line detection of coal quality components is of great significance.Because it not only can simultaneously detect many kinds of elements but also sample preparation is simple.It has been applied to the field of coal quality testing,however,there is a great distance with the actual application of power plant online.In order to accelerate the practical application of LIBS coal quality online detection technology,the research contents and results are as follows: from the characteristics of LIBS and the characteristics of the granular flow of pulverized coal:1)Taking into the actual industrial furnace pulverized coal status,an experimental platform is built to simulate the pneumatic conveying,the main experimental device is introduced in defails.2)According to the LIBS spectrum of the pulverized coal particle flow,the spectral data are pretreated.Firstly,the invalid spectrum,abnormal large and unusually small data filter out.Then the genetic algorithm is applied to reduce the interference of spectral line overlap.In the experiment,the 297.55 nm,280.21 nm and 285.1nm are obtained by decomposing and fitting the overlapping spectra of the 278~288nm band.Finally,based on the principle of the spectral line,the elements belonging to the above characteristic lines are identified,and the elements are Mg I,Mg II and Si.3)The influence of pulverized coal characteristics and the quantitative on the analysis results has been got.The model of the intensity of the element characteristic spectrum and the particle size of pulverized coal is established.The smaller the particle size is,the higher plasma temperature and electron density are obtained.The multiple linear regression analysis models are established,and the quantitative analysis is made to the industrial index of the coal quality components.The results shows that the values of ash,volatile and fixed carbon are 0.988,0.961 and 0.969,respectively,and RMSE are 0.085%,0.092% and 0.12% respectively.It can be seen that the use of multiple regression analysis is feasible for the quantitative analysis of coal powder industry index.Research shows that the accuracy of LIBS is improved.LIBS technology is used to detect the quality of coal quality,and provide the basis for real time on-line analysis.