张一鸣，葛邦国，高 玲，和法涛，吕绪强^*

（中华全国供销合作总社济南果品研究院，山东济南 250014）

摘要：以苹果片为原料，干燥过程中苹果片收缩比为指标，在单因素试验基础上，进行响应面优化试验，通过反向传播（Back propagation，BP）神经网络，对优化试验结果进行模拟分析并建立预测模型，结果表明：通过BP神经网络建立的苹果片干燥过程形态变化预测模型，迭代447次后得到最优模型，该模型R＝0.999，随机检测误差率均低于5%，对于苹果片干燥过程形态变化预测性较好。

关键词：苹果片干燥；神经网络；预测模型

中图分类号：TS255.42    文献标识码：A    文章编号：1674-506X（2022）04-0099-0005

Establishment of Morphological Change Prediction Model of Apple Slices During Drying Based on Neural Network

ZHANG Yiming, GE Bangguo, GAO Ling, HE Fatao, LV Xuqiang^*

（Jinan Fruit Research Institute All China Federation of Supply and Marketing Co-operatives, Jinan Shandong 250014, China）

Abstract：Taking apple slices as raw materials and taking the shrinkage ratio of apple slices during drying as the index, the response surface optimization test is carried out on the basis of single factor test. The optimization test results are simulated analysis and the prediction model is established by back propagation neural network. The results showed that the prediction model of morphological change of apple slices during drying established by BP neural network, the optimal model is obtained after 447 iterations. The model R＝0.999, and the random detection error rate is less than 5%. It has good predictability for the morphological changes of apple slices in the drying process.

Keywords：apple slice drying; neural network; prediction model

doi：10.3969/j.issn.1674-506X.2022.04-016

收稿日期：2021-12-01

基金项目：山东省重点研发计划项目（2019GNC106019）；高端外国专家引进计划（GL20200124001）；国家重点研发计划（2017YFD0400905-3）

作者简介：张一鸣（1993-），男，硕士，实习研究员。研究方向：果蔬干燥及成分评价。

*通信作者：吕绪强（1988-），男，本科，助理研究员。研究方向：新型果蔬加工制品研究。

引用格式：张一鸣,葛邦国,高玲,等.基于神经网络的苹果片干燥过程形态变化预测模型建立[J].食品与发酵科技,2022,58(4):99-103.