论文部分内容阅读
以甲苯、氯苯和正庚烷混和三组分体系为分析样品，应用模型转移法来校正不同仪器和不同测量时间导致的近红外光谱之间的差异。在校正测量时间导致的差异时，以16日由 Bruker 近红外光谱仪扫描的光谱作为主光谱，而17日由同台仪器扫描的光谱作为从光谱。在校正仪器导致的光谱差异时，以16日由 Bruker 近红外光谱仪扫描的光谱作为主光谱，而在同一天由 Perkin-Elmer 近红外光谱仪扫描的光谱作为从光谱。采集光谱后，用分段直接校正(PDS)和典型相关分析(CCA)校正从光谱。结果显示：虽然仪器和测量时间差别均导致从光谱偏离主光谱，但是 CCA 的效果更佳，使得校正后的从光谱能用主光谱的模型准确预测样品含量。“,”Calibration transfer methods were used to correct near infrared spectra (NIR) of tri-component mixtures including methylbenzene, chlorobenzene and n-heptane for their variations resulting from differences of measurement date and instrument. For correcting spectra between different measurement dates and instruments the spectra of mixtures measured in spectrometer of Bruker on date 16 were assigned as primary spectra while the spectra of same mixtures detected in spectrometer of Bruker on date 17 and in spectrometer of Perkin-Elmer on date 16 were set as secondary spectra for changes of measurement date and instrument respectively. After spectra obtained, two calibration transfer methods including canonical correlation analysis (CCA) and piecewise direct standardization (PDS) were used to correct the secondary spectra. The results showed that although variations of both measurement date and instrument can cause secondary spectra to deviate from primary spectra, the CCA can correct the secondary spectra to use the model of primary spectra to predict those three components with high accuracy.