文章编号:1673-887X(2023)05-0051-04基于BP神经网络的智慧农业云服务平台设计褚喆(河南测绘职业学院,河南郑州450015)摘要传统云服务平台在处理大量数据时准确率较低,为此,设计基于BP神经网络的智慧农业云服务平台。总体架构方面,采用PHP、JavaScript等技术相结合的方式,将整体架构分为3个不同层级,实现平台内部层级的动态交互;系统软件设计方面,以D-S证据理论算法为基础,设计农业信息采集模块;建立BP神经网络模型,寻找权重数值与固定阈值的最优值,确定模型拟合状态的精准度;采取B/S与C/A/S结构设计介质库管理模块,利用IP网络平台实现数据的传输与备份。测试结果表明,所设计的云服务平台,当信息数量增加到6000个时,系统的平均准确率为95.17%,可达到预期效果。关键词BP神经网络;智慧农业;云服务;平台构建中图分类号S126文献标志码Adoi:10.3969/j.issn.1673-887X.2023.05.019DesignofIntelligentAgriculturalAutomationMonitoringPlatformBasedonBPNeuralNetworkChuZhe(HenanCollegeofSurveyingandMapping,Zhengzhou450015,Henan,China)AbstractAbstract:Theaccuracyoftraditionalcloudserviceplatformislowwhenprocessinglargeamountsofdata.Therefore,theintelligentagriculturecloudserviceplatformbasedonBPneuralnetworkisdesigned.Intermsofoverallarchitecturedesign,PHP,JavaScriptandothertechnologiesarecombinedtodividetheoverallarchitectureintothreedifferentlevelstorealizethedynamicinteractionoftheinternallevelsoftheplatform.Intheaspectofsystemsoftwaredesign,agriculturalinformationacquisitionmoduleisdesignedbasedonD-Sevidencetheoryalgorithm.BPneuralnetworkmodelwasestablishedtofindtheoptimalvalueofweightvalueandfixedthresholdvalue,andtodeterminetheaccuracyofmodelfittingstate.B/SandC/A/Sstructureareadoptedtodesignmediali‐brarymanagementmodule,andIPnetworkplatformisusedtorealizedatatransmissionandbackup.Thetestresultsshowthatthede‐signedcloudserviceplatform,whenthenumberofinformationincreasesto6000,theaverageaccuracyofthesystemis95.17%,whichcanachievetheexpectedeffect.Keywords:BPneuralnetwork,intelligentagriculture,cloudservices,platformconstruction随着互联网与物联网的不断发展,很多网络应用开始提供云服务平台。目前的云服务平台划分为...