Research On Coal Quality Detection Method Of Pulverized Coal Using Laser-induced Breakdown Spectroscopy Under Physical Constraint

In industries such as coal-fired power generation,the real-time,rapid,highly sensitive acquisition of coal quality information can provide strong guidance for the production processes of coal-fired power plants.However,the existing off-line coal quality detection method based on thermogravimetric analysis is complicated and time-consuming,the severely lagging detection results can't meet the need for fast detection in the industrial field.Laser-induced breakdown spectroscopy(LIBS)is an emerging spectroscopy technology with many advantages such as rapid detection,high sensitivity,online and multi-element simultaneous detection,it has a wide application prospect in the field of coal quality detection.However,being limited by suitable coal sample preparation method and serious matrix effect interference,the industrial application of LIBS for coal quality online detection is still immature.Because of the above reasons,the physical constraint method was applied to prepare sample well in this paper,and the coal quality detection method for pulverized coal by LIBS was studied,the aim is to improve the online detection capability of LIBS technology applied to the industrial field.The detailed contents are as follows:(1)A simple sample preparation device based on physical constraint was designed and manufactured,and the preliminary experiments were carried out on the detection system composed of the device and the LIBS module.The situation of pulverized coal preparation by the device was specifically explored and the optimal sample amount of 15 g was determined.The structure of the detection system was improved,and the optimization of the main experimental parameters was completed.The results showed that the physical constraint device can prepare pulverized coal samples well,and the detection system was in a better state.(2)Furthermore,the spectral characteristics and quantitative performance of coal quality detection under physical constraint were studied by comparing with other detection forms.It was found that physical constraint can form a spatial constraint on the plasma so that the intensity and signal-to-noise ratio of the characteristic spectra related to the ash were enhanced by 2 to 3 times.Besides,the physical confinement effect can also effectively avoid sample ejection.The RSD was reduced to 7.75%,significantly improving the stability of the spectra.To verify the quantitative performance,partial least square regression was used to establish a quantitative detection model for ash and volatile matter under physical constraints.The predicted average relative error and root mean square error are 4.14%,1.2wt.% and 3.46%,0.34 wt.%,respectively.That showed that while the detection efficiency is greatly improved,its detection accuracy has reached the same level as that of conventional coal pellet modelling.(3)Aiming at the problem that the matrix effect seriously affects the accuracy of quantitative detection,an improved internal standard method is proposed to process the spectral data,achieving the purpose of improving the quantitative detection performance of coal quality by LIBS under physical constraints.For this method,the spectral line intensity and mass fraction of carbon were used together to correct the original spectra,which enhanced the correlation between spectral line intensity and industrial detection indicators to more than 20%.The artificial neural network prediction model of ash content,volatile content and calorific value was established on the processed spectra,achieving a huge improvement in detection accuracy of 79%,35% and 53% respectively and proving the efficiency of the improved internal standard methodIn summary,a systematic study on the coal quality detection method of pulverized coal by LIBS under physical constraints was carried on.With physical constraint,the rapid preparation and accurate detection of pulverized coal were realized,and the use of improved internal standard processing has greatly improved the quantitative accuracy of the detection system,laying a certain foundation for the application of coal quality online detection by LIBS.