利用卡尔曼滤波算法的粮情预测与分析  

Research on grain situation prediction based on Kalman filter algorithm

作　　者：刘帅 尹强 黄强 张永林 刘晓鹏 LIU Shuai;YIN Qiang;HUANG Qiang;ZHANG Yonglin;LIU Xiaopeng(School of Mechanical Engineering,Wuhan Polytechnic University,Wuhan 430023,China)

机构地区：[1]武汉轻工大学机械工程学院,武汉430023

出　　处：《武汉轻工大学学报》2024年第2期78-84,共7页Journal of Wuhan Polytechnic University

基　　金：湖北省科技厅2023年科技人才服务企业项目(编号:2023DJC105);湖北省技术创新计划重点研发专项项目(编号:2023BBB018);2023年武汉轻工大学科研项目(编号:2023Y32).

摘　　要：粮食的安全存储,通常需要对温度、湿度、水分和CO_(2)浓度等环境信息进行常态化采集并判断粮食是否处于安全状态,为了减轻管理人员的劳动量以及做到及时处理,本文利用优化后的卡尔曼滤波算法,构建出粮情预测模型对粮库环境信息进行预测,达到提前预知粮库环境信息的变化情况。并将所构建的粮情预测模型与指数平滑法模型进行对比,结果表明:相较于指数平滑法模型,卡尔曼滤波算法所构建的粮情预测模型对各环境信息的平均误差较小,用作预测粮库环境信息行之有效,有一定的实用价值。To ensure the safety of stored grain,it's essential to regularly collect environmental data such as temperature,humidity,moisture,and CO2 levels and determine whether the grain is safe.Previously,this process required manual labor from grain storage managers,which was both time-consuming and prone to human error.This could even result in unnecessary loss of grain due to delayed detection of abnormal grain conditions.To reduce labor and achieve timely processing,this paper proposes using an optimized Kalman filter algorithm to construct a grain prediction model.The model predicts the environmental conditions of the grain storage,enabling managers to anticipate changes in these conditions in advance.The grain prediction model is compared with the exponential smoothing model,and the results show that the former has a smaller average error for each environmental variable.This makes it an effective and practical tool for predicting environmental conditions in grain storage facilities.

关 键 词：粮情预测 卡尔曼滤波算法 粮食储存 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]