基于近红外光谱的食用油酸价和过氧化值自动化检测

Automatic Determination of Acid Value and Peroxide Value of Edible Oils by Near-infrared Spectroscopy

为实现食用油酸价和过氧化值的自动化快速检测,以常见食用油为材料,利用近红外自动分析仪结合连续进样流通池,建立食用油酸价和过氧化值的定量模型并进行模型验证。结果表明:在波数范围为5 500～4 600 cm-1时,光谱预处理选择(6 524,4 823)两点基线校正和标准正交变换,酸价定量模型验证相关系数和预测标准偏差分别为0.987 3和0.114 mg/g;在波数范围为6 050～4 450 cm-1时,选择一阶求导和Norris导数平滑,过氧化值定量模型验证相关系数和预测标准偏差分别为0.995 8和0.90 mmol/kg;模型盲样验证效果良好,模型可行,通过近红外自动分析仪每小时可检测90个样品。

For automatic determination of acid value and peroxide value of edible oils, using edible oils as raw materials, calibrations were developed and validated for determination of acid value and peroxide value by near-infrared spectrometer with continuous samples injection cell. The correlation coefficient R and the root mean square errors of prediction （RMSEP） of the calibration for acid value were 0. 987 3 and 0. 114 mg/g in range of 5 500 - 4 600 cm - 1 with baseline correction point at 6 524 cm- 1 and 4 823 cm - and standard normal variant （SNV）. The R and RMSEP for peroxide value of calibration were 0. 995 8 and 0.90 mmol/kg in range of 6 050 - 4 450 cm-1 with first derivative and Norris filter. The results of blind samples validation were good, the calibrations were workable, throughput of 90 samples can be achieved per hour.