基于低场核磁共振技术构建韧性饼干中水分含量无损定量预测模型

Construction of non-destructive quantitative prediction model of the moisture content of tough biscuits based on low-field nuclear magnetic resonance

中文关键词:  韧性饼干  低场核磁共振技术  水分  快速检测

英文关键词:tough biscuits  low-field nuclear magnetic resonance  moisture  quick detection

中文摘要:

      目的 基于低场核磁共振技术(low-field nuclear magnetic resonance, LF-NMR)弛豫特性建立韧性饼干水分含量的预测模型, 探讨LF-NMR无损快速预测韧性饼干水分含量的可行性。方法 采用LF-NMR对韧性饼干的CMPG (Carr-Purcell-Meiboom-Gill)序列信号进行采集, 利用直接干燥法检测韧性饼干中水分的实际含量, 建立多元回归分析、K最邻值回归分析(K-nearest neighbor, KNN)和高斯过程回归(Gaussian process regression, GPR)模型。结果 3种模型中KNN回归模型预测集决定系数最大为0.9932, 均方根误差最小为0.2542, 因此韧性饼干中水分与低场核磁弛豫特性的分析中, KNN回归分析的模型最优。结论 使用LF-NMR分析仪预测韧性饼干中水分含量是一种无损、快速的方法, 对于韧性饼干样品中水分快速预测建模结果良好, 因此利用低场核磁共振技术快速预测韧性饼干样品中的水分含量方法可行。

英文摘要:

      Objective To construct the prediction model of the moisture content of tough biscuits based on the relaxation characteristics of low-field nuclear magnetic resonance (LF-NMR), and explore whether it is feasibility of LF-NMR for nondestructive and rapid prediction of moisture content of tough biscuits. Methods The CMPG (Carr-Purcell- Meiboom-Gill) sequence signals of tough biscuits were collect by LF-NMR, and the actual content of the moisture in the tough biscuits was detected by the direct drying method, the multiple regression analysis, K-nearest neighbor (KNN) regression and Gaussian process regression (GPR) models were established, to explore the LF-NMR in measuring moisture content of tough biscuits. Results The predictive set of KNN regression models in the 3 kinds of models had the largest decision coefficient of 0.9932, the smallest average square root error was 0.2542, the model of KNN regression analysis of moisture and LF-NMR relaxation characteristics in tough biscuits was optimal. Conclusion LF-NMR is a nondestructive and fast detection method for the moisture content prediction in tough biscuits, the rapid moisture detection models in the same type of tough biscuits samples are good, therefore, it is feasible to quickly predict the moisture content in tough biscuits samples by LF-NMR.