Research On Soybean Optimum Cultivation Management Decision Support System Based On Struts

With the popularity and development of internet technology, the single-machine of the expert system is no longer suited to network application environment. It is a major trend that the expert system launched the internet version has been researched.The paper analyzes the application of the current computer technology in the agricultural expert system and the network infrastructure, and discusses the theory and technology on constructing crops cultivation agricultural decision support system. The study is based on"browser /Tomcat server /Oracle database"three-tier network architecture model, uses framework struts Java Web MVC. The soybean cultivation and management optimization decision support system is researched and developed with the actual soybean production in Inner Mongolian (such as hohhot, tongliao ,etc). The separation of view layer, logical service layer and data layer is good achieved in the system. Thereby, the system's maintainability and code reuse rate are enhanced.According to the characteristics of soybean cultivation areas, the knowledge data representation, combining with the pattern representation and the knowledge unit representation, is used in the system. Under a cross-platform system compeleted by using the module technology and the distribution network technology, this method is helpful to the inference machine's design and implementation. The entire inference engine is divided into seven independent Beans, based on the forward and backward inference strategy. Each Bean mutually cooperates to complete reasoning. It is an alternative to single-thread/single module inference engines and improves the speed of reasoning. BP neural networks realized by JAVA programming provide predicted values based on yield for the target yield optimization. The interaction of it and the entire system is achieved through Java technology. Finally, the paper shows the system's realization.