乳化炸药水相组分与析晶点的近红外光谱检测技术OA

The main components of the aqueous phase of emulsion explosives include ammonium nitrate NH4 NO3,sodium nitrate NaNO3,and water.A multi-component quantitative analysis model for near-infrared spectroscopy based on partial least squares metho was proposed.According to the spectral-physicochemical value c symbiosis distance classifica-tion method,the infrared spectral data of the aqueous phase of emulsion explosives were divided into samples to obtain a calibration set and a prediction set.Standard normal variable transformation(SNV)+Savitzy-Golay(S-G)convolution smoothing+multivariate scattering correction(MSC)+S-G convolution differentiation preprocessing method were used to extract the features of the correlation spectral intervals,and a partial least squares regression(PLS)prediction model for the aqueous phase component content of emulsion explosives was established.In the established model,the determination coefficient R2C of the calibration set is 0.994,and the root mean square error RMSEC is 1.600 2.The determination coeffi-cient R2P of the prediction set is 0.952,and the root mean square error RMSEP is 1.700 9.The relative analysis error RPD is 3.324 9.The results indicate that the accuracy,stability,and reliability of the quantitative analysis model meet practical needs.Near-infrared spectroscopy can achieve online detection of the main component content of emulsion explosives.Meanwhile,the near-infrared spectroscopy analysis model was used to measure the ammonium nitrate crystallization point in the aqueous phase.The average standard deviation is less than 0.3,indicating better stability and higher accuracy com-pared to manual measurement.This model was also used to investigate the influence of the proportion of aqueous phase components on the stability of crystallization points.It is concluded that the optimal sodium nitrate content ratio in aqueous phase is about 4.0%.