基于NIR光谱的固体推进剂代用料混合特性

发布时间：2018-04-25 11:07

本文选题：近红外光谱 + 固体推进剂　；参考：《固体火箭技术》2016年03期

【摘要】：建立了AOTF近红外光谱仪检测固体推进剂中多固体组分混合的方法,以PVB、MA、Al粉三固体组分在水中的混合为基础,扫描样品得到近红外光谱,经平滑、均值归一化、标准归一化(SNV)、基线校正和一阶、二阶导数预处理后,用偏最小二乘法(PLS)、交互验证建立3种组分的定量校正模型。结果表明,不同预处理方法对结果影响不一,其中使用SNV预处理消除了光谱散射对结果的影响,得到的各组分定量校正模型最优。3种组分中,定量校正模型决定系数均在0.95以上,以MA的模型决定系数最大,为0.991 377,对近红外光谱本身没有特殊吸收的组分Al粉,也得到了很好的定量校正模型,模型决定系数达到0.984 637。3种组分的定量校正模型对未知样品进行预测时,绝对误差均小于1%。因此,可用近红外光谱分析法对多固体组分的混合进行测试。
[Abstract]:A AOTF near-infrared spectrometer was established to detect the mixing of multi-solid components in solid propellants. Based on the mixing of three solid components in water, the near infrared spectra were obtained by scanning the samples, which were smoothed and normalized by the mean value. After standard normalized SNV, baseline correction and first order and second order derivative pretreatment, partial least square method was used to establish the quantitative calibration model of three components. The results show that different pretreatment methods have different effects on the results, and the effects of spectral scattering on the results are eliminated by using SNV pretreatment. The determination coefficients of quantitative calibration models are all above 0.95, and the determination coefficient of MA model is the largest, which is 0.991. For Al powder, which has no special absorption in NIR spectrum itself, a good quantitative correction model has been obtained. In the prediction of unknown samples, the absolute error is less than 1 when the quantitative calibration model with a determination coefficient of 0.984 637.3 components is used to predict the unknown samples. Therefore, the mixing of multi-solid components can be measured by near-infrared spectroscopy.
【作者单位】：南京理工大学化工学院;山西北方兴安化学工业有限公司;
【基金】：基础产品创新计划火炸药科研专项、江苏高校优势学科建设工程资助项目(PAPD)
【分类号】：V512

【相似文献】