BP神经网络结合有效积温预测速冻水饺变温冷藏货架期  

Prediction of shelf life for quick-frozen dumpling based on BP neural network and effective accumulated temperature theory

作　　者：潘治利[1] 黄忠民[1,2] 王娜[1] 张卓[1] 谢新华[1] 索标[1] 卢邦贵[1] 艾志录[1] 

机构地区：[1]河南农业大学食品科学技术学院,郑州450002 [2]郑州市三全食品股份有限公司,郑州450000

出　　处：《农业工程学报》2012年第22期276-281,共6页Transactions of the Chinese Society of Agricultural Engineering

基　　金：河南省科技厅重点科技攻关计划项目(编号:082102320006);国家自然科学基金资助项目(31071598);郑州市科技局重大科技攻关项目(072SGZN12029);国家"十二五"科技支撑计划(2012BAD37B06-04)

摘　　要：为了解决速冻食品在温度波动贮藏过程中的货架期预测问题,准确监测其品质变化趋势,该文以速冻水饺为研究对象,将其在-28℃～-12℃进行冷藏,测定酸价、过氧化值、饺皮水分含量、亨特白度等理化指标,并结合感官评价与有效积温理论,应用BP神经网络技术预测速冻水饺的货架期。并与动力学模型预测结果进行比较。结果表明,测试集样本的距货架期终点积温的预测值与实际测定值拟合度较好,最大误差为3.29%,模型验证最大误差为2.74%。BP模型的距货架期终点时间的最大误差为3.45%低于传统动力学模型预测的误差(5.62%)。BP神经网络预测模型为速冻食品货架期预测提供了一种新途径。In order to well predict the shelf life of quick frozen food during the temperature-fluctuation storage, and accurately monitor its quality changes, frozen-dumplings stored with fluctuant temperature from -28 to -12℃ were studies. Physicochemical indices including acid value, peroxide value, moisture content of dumpling skin, brightness of dumpling skin and sensory evaluation were determined. BP neural network was applied to predict the shelf life of quick-frozen dumplings combined with effective accumulated temperature theory. And kinetic model was used to take comparative analysis. The results showed that the predictive value for BP neural network fitted well with the experimental value, and the maximum error was 3.29%. The maximum error of the test experiment for BP neural network was 2.74%, which was less than that of the kinetic model （5.62%）. BP neural network provides a new way to predict the shelf life of quick frozen food.

关 键 词：贮藏 温度 品质调控 速冻水饺 货架期 BP神经网络 

分 类 号：TS205.7[轻工技术与工程—食品科学]