节点文献

神经网络测报软件的设计实现

Design and Implementation of the Neural Network Forecasting System

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 秦淑莲姜玉英曾娟张孝羲

【Author】 QIN Shulian1,JIANG Yuying2,ZENG Juan2,ZHANG Xiaoxi3(1.College of Science and Information Engineering,Qingdao Agricultural University,Qingdao,266019,China;2.Pest Forecasting Division,National Agro-Technique Extension and Service Center;3.Key Laboratory of Pest Monitoring and Management of Chinese Agricultural Ministry,Department of Entomology,Nanjing Agricultural University)

【机构】 青岛农业大学理学与信息科学学院全国农业技术推广服务中心测报处南京农业大学昆虫学系,农业部病虫监测与治理重点实验室

【摘要】 用Visual FoxPro数据库管理系统管理数据,用ADO控件连接数据库,用Visual Basic编写程序代码,完成了神经网络测报模型,作为病虫害测报系统软件的一个功能模块,该模块包括六个功能子模块:数据导入子模块用于将数据从磁盘文件中导入系统;参数设置子模块用于对各种初始参数进行设置;数据标准化处理子模块用于对原始数据进行标准化处理;激活函数的选择子模块用于选择不同的激活函数;学习训练子模块用于对模型进行学习训练;测报子模块用于预测预报病虫害的发生情况。为病虫害的预测预报提供了一个操作简单、方便实用的工具软件平台。测报工作者可以根据自己实际的原始数据组建成当地实用的神经网络模型,并测报应用。

【Abstract】 The neural network forecasting model,as a function module of the pest and disease forecasting system software,was designed and implemented by Visual Basic,using Visual FoxPro database management system to manage data,with the ADO control objects to connect to the database.There were six functional sub-modules in the module,such as: Data inputting sub-module was used to input data from a disk file;Parameter setting sub-module was used to set various initial parameters;Data Standardization processing sub-module was used to standardize the Original data;Active function Selecting sub-module was used to select a different active function;Learn training sub-module was used to train learn for real model;Forecasting sub-module was used to predict and forecast the occurrence of pests and diseases.Prediction of workers may use this platform to set up the real neural network forecasting model and release the real occurrence of pests and diseases.It provided a simple,convenient and practical tools software for pests and diseases forecasting.

【基金】 公益性行业(气象)科研专项经费项目“农作物病虫害发生气象条件监测、预警和评价技术”(GYHY201006026)
  • 【文献出处】 青岛农业大学学报(自然科学版) ,Journal of Qingdao Agricultural University(Natural Science) , 编辑部邮箱 ,2013年01期
  • 【分类号】TP183;TP311.52
  • 【被引频次】2
  • 【下载频次】37
节点文献中: 

本文链接的文献网络图示:

本文的引文网络