The Research Of Expert System Based On Web For Intelligent Management Of Muskmelon In Greenhouse

With the development and popularization of Internet, People can more easily gain information using a browser. Most of traditional muskmelon expert systems are off-line mode. It does not facilitate the distribution and update of the system. But the Web based systems are lock of support for development model, and then they simulate the dynamic procedure of growth.In this paper, we introduce the concepts and characteristic of expert system. After comparing the popular expert system in agriculture, we point out the trend of existing system and show the difficulty which we are facing. We propose a method using Java Web and Drools, so we can use Drools to be our inference engine and manage muskmelon culture knowledge as DSL(Domain Specific Language).We have chosen several development models and try to get the parameters which we can use to run the models. We use the chosen models to simulate the development procedure of muskmelon, and results are transfer to the expert system which will give a readable explain about how to manage the muskmelon.Then this paper presents the architecture of the whole system. It is a MVC-based Java Web application. And Jquery is used for presentation of Web pages and asynchronous communication with server. In the server we use Struts2 to deal and dispatch the HTTP-request. SQLite3 is used to be the database to store the environment data and other information.We have developed a muskmelon management software to provide the functions such as prediction of crop growth stages and muskmelon LAI development simulation. It makes full use of open source software to cut down the costing.