基于气流-激光检测技术的面包老化表征

Characterization of bread staling based on airflow-laser detection technology

中文关键词:  伯格斯模型  蠕变  面包  老化  气流-激光  定量模型  快速检测

英文关键词:Burgers model  creep  bread  staling  airflow-laser  quantitative model  rapid detection

中文摘要:

      目的 探索应用气流-激光检测技术实现面包老化快速定量检测方法。方法 使用基于气流-激光检测技术的检测装置进行蠕变测试对面包黏弹性参数进行采集, 分别使用采集到的蠕变阶段全参数和基于伯格斯模型提取的黏弹性参数, 结合不同预处理方法建立基于水分含量的水分损失速率和基于硬度的老化率的多元线性回归分析和偏最小二乘回归分析模型。结果 使用蠕变阶段全参数建立的预测模型取得最佳预测效果, 对于老化率模型, 使用卷积平滑(savitzky-golay, S-G)结合偏最小二乘回归最佳建模结果为校正集和验证集相关系数分别为0.971和0.959, 校正集均方根误差和验证集均方根误差分别为9.723和10.721; 对于水分损失速率模型, 使用1阶导加S-G平滑结合偏最小二乘回归最佳建模结果为校正集和验证集相关系数分别为0.984和0.968, 校正集均方根误差和验证集均方根误差分别为0.002和0.002。结论 使用气流-激光检测技术可以对面包老化进行快速、简单、可靠的表征, 实现对面包老化的定量检测。

英文摘要:

      Objective To explore the method for the application of airflow-laser detection technology to achieve rapid quantitative detection of bread staling. Methods The creep test was performed using a detection device based on airflow-laser detection technology to collect the bread viscoelastic parameters. The full parameters of the creep phase and the mechanical properties parameters extracted based on the Burgers model were collected and combined with different pretreatment methods to establish multiple linear regression analysis and partial least squares regression analysis models for the moisture loss rate based on moisture content and staling rate based on hardness. Results The prediction model established by using the full parameters of creep stage achieved the best prediction effect. For the staling rate model, the best modeling results using savitzky-golay (S-G) convolutional smoothing combined with partial least squares regression were 0.971 and 0.959 for the calibration set and validation set correlation coefficients, respectively, and 9.723 and 10.721 for the root mean square error of the calibration set and validation set, respectively. For the moisture loss rate model, the best modeling results were 0.984 and 0.968 for the calibration set and the validation set, respectively, and 0.002 and 0.002 for the root mean square error of the calibration set and the root mean square error of the validation set, respectively, using the first-order derivative plus S-G convolutional smoothing combined with partial least squares regression. Conclusion The use of airflow-laser detection technique can provide a fast, simple and reliable characterization of bread staling, and realize the quantitative detection of bread staling.