基于BP神经网络的仓储小麦的生理生化指标预测  

Prediction of Physiological and Biochemical Indexes of Stored Wheat Based on BP neural Network

作　　者：张爱琳 梁筱妍 杨薇 徐学彬 Zhang Ailin;Liang Xiaoyan;Yang Wei;Xu Xuebin(Tianjin Agricultural University College of Food Science and Biotechnology,Tianjin 300384;Tianjin Agricultural University College of Basic Science,Tianjin 300384;Tianjin Jinghai Ancient City Grain Reserve Co.,Ltd.,Tianjin 301604)

机构地区：[1]天津农学院食品科学与生物工程学院,天津300384 [2]天津农学院基础科学学院,天津300384 [3]天津市静海古城粮食储备有限公司,天津301604

出　　处：《中国粮油学报》2023年第2期153-158,共6页Journal of the Chinese Cereals and Oils Association

基　　金：国家重点研究与发展计划(2019YFC1605305);天津市重点研发计划项目(20YFZCSN00300);天津市农业科技成果转化与推广项目(202101080)。

摘　　要：BP神经网络算法在粮食仓储领域拥有巨大的应用价值和潜力。本研究尝试将BP神经网络引入仓储小麦品质预警模型,以天津储粮抽检数据为对象,通过对室内温度的记录,样品水分、淀粉、蛋白质等11项生理生化指标的定期检测,利用BP神经网络算法进行仓储小麦的品质预测与影响分析。仿真结果表明,基于BP神经网络的数据预测方法具有较小的过程误差和较高的结果准确性,为仓储小麦的品质预测提供了一种有效的研究方法。BP neural network algorithm has great application value and potential in the field of food storage. In the present study, BP neural network was introduced into the quality warning model of stored wheat, and the quality prediction and influence analysis of stored wheat were carried out by recording the indoor temperature and regularly detecting elevenphysiological and biochemical indexes, such as moisture, starch and protein, based on the sampling data of stored grain in Tianjin. The simulation results showed that the data prediction method based on BP neural network has less process error and higher result accuracy, providing an effective research method for the quality prediction of stored wheat.

关 键 词：小麦仓储 神经网络 品质劣变 粮仓预测 

分 类 号：TS212.2[轻工技术与工程—粮食、油脂及植物蛋白工程]