基于径向基函数网络与WebGlS的茶叶病害预测

摘要

By using mode integration of GIS spatial analysis and MATLAB numerical functions, the system described in this article applies MATLAB neural network toolbox to call parameters fi'om the GIS server to establish RBF neural network prediction model, and then send the content back to the server. It makes a seamless connection between GIS and the predictive models and shows in real-time dynamic forecasting model by the WebGIS, which achieve the forecast tea disease for querying, forecasting and diagnostics. In addition, the system has a processing and analysis function in spatial and attribute data and transform abstract data into clear and concise electronic maps in graph or table format to show tea disease type and the rule of geographic distribution. It makes efficient service for the disease diagnosis and prevention conveniently.%利用GIS空间分析与MATLAB数值计算功能的进行模式整合，应用MATLAB神经网络工具箱，从GIS服务器调用参数信息建立RBF神经网络预测模型，将内容反馈给服务器，形成GIS与预测模型的无缝连接，通过We—bGIS对预测模型进行实时动态显示，实现对茶叶病害的预测、查询和诊断。此外，系统具有对空间、属性数据的分析与处理功能，并将抽象的数据转化成清晰简明的电子地图，以直观的图、袁显示病害类型及病害地域分布规律，为茶叶病害的诊断防治提供快捷、高效的服务。