基于BP神经网络的智慧农业云服务平台设计

Design of Intelligent Agricultural Automation Monitoring Platform Based on BP Neural Network

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。

The accuracy of traditional cloud service platform is low when processing large amounts of data.Therefore,the intelligent agriculture cloud service platform based on BP neural network is designed.In terms of overall architecture design,PHP,JavaScript and other technologies are combined to divide the overall architecture into three different levels to realize the dynamic interaction of the internal levels of the platform.In the aspect of system software design,agricultural information acquisition module is designed based on D-S evidence theory algorithm.BP neural network model was established to find the optimal value of weight value and fixed threshold value,and to determine the accuracy of model fitting state.B/S and C/A/S structure are adopted to design media li‐brary management module,and IP network platform is used to realize data transmission and backup.The test results show that the designed cloud service platform,when the number of information increases to 6000,the average accuracy of the system is 95.17%,which can achieve the expected effect.