粮堆温度预测插值算法的研究  

Prediction and Interpolation Algorithm of Temperature of Grain Bulk

作　　者：张明跃 周慧玲[1] 钱荣荣[1] 汪中明 刘尚峰 Zhang Mingyue;Zhou Huiling;Qian Rongrong;Wang Zhongming;Liu Shangfeng(College of Artificial Intelligence, Beijing University of Posts and telecommunications, Beijing 100089;Academy of National Foof and Strategic Reserves Administration, Beijing 100037;Anhui Modern Grain Logistics Center, Shucheng 231323)

机构地区：[1]北京邮电大学人工智能学院,北京100089 [2]国家粮食和物资储备局科学研究院,北京100037 [3]安徽现代粮食物流中心库,舒城231323

出　　处：《中国粮油学报》2021年第12期136-143,共8页Journal of the Chinese Cereals and Oils Association

摘　　要：温度变化对粮堆内的微生物滋长和虫害活动有着十分重要的影响,对粮堆温度进行准确监测和预测是保障储粮品质和安全的关键因素。本文通过对实际粮堆温度传感器节点阵列的历史数据进行挖掘分析,提出了基于时间-空间相结合的粮温预测模型,可对各个节点未来时刻的温度值进行预测,并可通过此模型对未知节点温度进行预测插值,以得到未来时刻粮库的温度云图。实验结果表明,与现在粮库常用的基于经验函数的温度预测算法和只考虑时间层面的机器学习预测算法相比,本文预测模型得到的节点温度预测均方误差分别降低了23.81和6.82,预测结果得到提升,并且通过预测插值得到的未来时刻粮堆温度场分布云图对粮库预知局部危险因素具有实际的指导意义。The temperature change of stored grain has the close relationship with the mould growth and insect pests activities in the grain bulk.Accurate monitoring and prediction of the temperature of grain bulk is the key factor to ensure the quality and safety of stored grain.In this paper,we proposed a grain temperature prediction model based on the historical temperatur sensor-nodes'data from the grain bulk,considering the sensor-nodes'timeless and spatiality together.This model can predict the temperature value of each node in the future,and get the temperature values of the unknown nodes,so as to obtain the temperature nephogram of grain bulk in the future.The experiment results show that,compared with the commonly temperature prediction algorithm based on empirical functions and prediction algorithm based on machine learning which only consider the time factor,the mean square error of node temperature prediction that obtained by ourproposed model is reduced by 23.81 and 6.82 respectively,and the prediction results are improved.Moreover,the future temperature distribution cloud map of grain bulk obtained by prediction interpolation can predict whether the grain situation is safe or not.This is very significant for guiding the practice.

关 键 词：粮堆温度 时序预测 空间插值 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]