Abstract: The real-time and accurate crop forecasting yield data is an indispensable constituent of agricultural production monitoring information. This paper introduces a crop yield forecasting system which has utilized Oracle DBMS as the database platform, adopted the system framework of Client/Server(C/S), and used Interactive Data Language(IDL) with the character of orienting matrix for the system development. The system integrates multiple models using different types of data together, implements visualization of model analysis, batch process, spatial query of GIS, and so on, and can realize the yield forecasting analysis of dominant food crops at the scale of provincial level to provide the agriculture and food departments with prompt data for decision-making analysis.

Key words: crop yield, multiple-model integration, database, forecasting system

摘要： 实时准确的农作物单产预测数据是粮食产量监测信息的重要组成部分。该文介绍了以大型Oracle数据库为数据管理支持平台的农作物单产预测系统。该系统采用客户端/服务器(C/S)的系统架构模式，使用面向矩阵的程序语言IDL。将使用不同类型数据的多个作物单产预测模型集成在一起，具有预测模型的可视化显示、相关批处理运算、结果的GIS空间查询等功能，可实现全国省级行政区域的主要粮食农作物的单产预测，为相关农业和粮食管理部门提供及时的决策支持信息。

关键词: 农作物单产, 多模型集成, 数据库, 预测系统

CLC Number: 

• TP391