Study And Implementation Of The Key Technologies Of The Rice Diseases And Pests Warning Based On WebGIS

Rice diseases and pests are among the most important factors of grain output. Nearly20years, rice diseases and insect pests have been seriously defending the rice development. Therefore, timely and accurate predictions play an important role in the development of the rice in growing area and sustainable development of regional rural economy.With the development of information technology, computer networks, data management and GIS technology has been widely applied to pest and disease detection and forecast. To adapt to crop pest data and information, networking, standardization requirements, rice diseases and pests warning system in Hunan province based on WebGIS has been established according to the Chiloe suppressalis data and meteorological data from1992to2003in the rice-growing region of Taoyuan. Through the simulation analysis,the prediction result was ideal and has some practical value.The researches of paper as follow:(1) The attribute data of rice diseases and insect pests have been reduced as variables of nonlinear regression warning model to predict the grade of rice diseases and insect pests by distinguishable matrix based on rough set theory, the predicted results was full compliance with the actual.(2) According to the characteristics of the occurrence of rice pest and disease, use BP neural network method to diagnose rice diseases and discriminate pest species.(3) In order to forecast how much diseases and pets would emerge, the mixed model of time series combining Trend Extrapolation with Exponential smoothing has been established, the prediction result was good.(4) In order to provide data for short-term prediction, Application Seasonal Exponential smoothing method has been used to establish the seasonal forecasting model.(5) In order to achieve space prediction, GIS Buffer Analysis has been used to show the region where diseases and pets could endanger.(6) WebGIS warning system construction. Using Arc IMS,PostgresSQL/GIS and IIS to build WebGIS platform. This system converted abstract data to the simple clear map and statistical form for user inquires convenient. Through predictive analytics,showing the spread area and dangerous severity of rice diseases and pests and provides a visual network service platform for prediction.