基于LSTM算法的玉米籽粒储藏温度预测

Corn kernel storage temperature prediction based on the LSTM algorithm

为减少储粮损失和虫霉等的发生,该文利用自制试验仓及检测系统检测储藏玉米籽粒不同位置的粮温,分析粮堆温度变化及整个粮堆热量的传递过程。试验结果表明,仓外环境温度对粮温影响较大,仓内粮食温度变化与仓外环境温度变化相比较为滞后,粮堆第一层2号位置温度在检测周期中一直处于较低状态,温度最高位置出现在第四层12号位置。基于粮堆温度变化分析,该文开展了基于长短时记忆网络(LSTM)算法的玉米籽粒储藏粮温预测研究。结果表明:(1)对比预测值与试验值可知,粮堆第一、二、三、四层测试集的粮温准确率分别为0.62、0.89、0.83、0.79;(2)位于粮堆第二层和第三层的预测结果精度较高,试验仓粮堆底层和顶层温度易受环境温度影响,粮堆热量交换速度快,温度变化迅速,导致第一层和第四层预测结果精度偏低。该研究可为粮食储藏温度预测研究提供新思路。

[Objective]Grain storage is an extremely complex process that is easily affected by the external environment.A temperature gradient is formed between the inside and outside of the grain pile,making the pile prone to rot and mildew.[Methods]To reduce the loss of grain and the occurrence of pests and mildew,this study uses a self-made test chamber and detection system to detect the grain temperature at different locations of corn grain storage and analyze the temperature change and the heat transfer process across the entire grain pile.Next,the prediction of corn grain storage temperature based on the LSTM algorithm is investigated by analyzing the corn grain storage test data.Based on the attributes and characteristics of the data,linear algebra,probability theory,and algorithm principles are applied to form a comprehensive decision and predict the corn grain storage temperature.[Results]The test results show that the atmospheric temperature outside the bin greatly influences the grain temperature,and the grain temperature in the bin falls behind that outside the bin as well as the ambient temperature.The temperature in the bottom No.2 position always remain slow during the detection period,while the highest temperature occurs in the top No.12 position.[Conclusions]Based on the analysis of grain pile temperature change,this study investigates the prediction of grain storage temperature based on the LSTM algorithm.Compared with the experimental values,the accuracy of the test sets of the first,second,third,and fourth layers is 0.62,0.89,0.83,and 0.79,respectively.Thus,the prediction results of the second and third layers of the grain pile have higher accuracy.The bottom and top layers of the grain pile in the test bin are easily affected by the ambient temperature,the heat exchange rate of the grain pile is fast,and the temperature changes rapidly,yielding less accurate prediction results for these layers.The research presented in this study provides new insights into the prediction of grain storage temperature.