论文部分内容阅读
采用红外光谱(FTIR)和偏最小二乘(PLS)法,建立了地沟油掺混比例的衰减全反射(ATR)定量分析模型。分别采集大豆油、地沟油和二者不同掺混比例的红外光谱图,应用TQ化学计量学分析软件,对不同光谱的采集方式,不同化学计量学处理模式进行了比较,并对光谱区域,光谱的预处理方法,主成分因子数进行筛选。依据预测效果确定了最佳的预测模型,其相关系数为0.99,定标均方差(RMSEC)为4.31、预测均方差(RMSEP)为5.55。本法可用于地沟油鉴别的初期检测。
FTIR and PLS methods were used to establish the ATR quantitative analysis model. We collected the infrared spectra of soybean oil, waste oil and different blend ratios of the two separately. The TQ chemometrics software was used to compare the different spectral acquisition modes and different stoichiometric treatment modes. The spectral region, The pretreatment method, the main component factor screening. The best prediction model was determined based on the prediction results, with a correlation coefficient of 0.99, a root mean square error of calibration (RMSEC) of 4.31 and a root mean square error of prediction (RMSEP) of 5.55. This method can be used for the initial detection of waste oil identification.