Research And Implementation Of Wheat Disease Diagnosis Expert System Based On BP Network

Through the analysis and study of traditional artificial wheat disease diagnosis, there exist conditions of low inefficient and high error rates, and then the pesticides cannot be correct used in time after the wheat get sick, so as to reduce the yield of wheat. To solve the above problems, this paper intends to use expert systems to realize wheat disease diagnosis. However, the traditional expert systems have the shortcomings of difficult knowledge acquisition, low self-learning ability and single logical reasoning. Therefore, this paper uses BP neural network combined with expert systems to overcome the shortcomings of traditional expert systems to some extent.The system is built on the standards-based java EE and 3-layer B/S structure intelligent diagnosis expert systems of major wheat diseases. This system realizes the diagnosis of major 10 wheat diseases, and maintenance and browse information of related knowledge of wheat. And it will enable users to access the detail information of diseases, and combine with the output diagnosis results of the system to take effective measures, which will provide strong support for wheat production.The main contents of this paper are as follows:(1) Construction of wheat diseases diagnosis model based on BP network. Aiming at the specific situation of wheat, the Inference mechanism of expert system is established by using t he combination of expert system and BP neural network technology. According to the diagnos tic parameters that user input, the wheat disease diagnosis model diagnoses intelligently and t hen outputs the results. And the correct rate of diagnosis can reach 80.5%.(2) Build domain knowledge base. First of all, obtaining the knowledge and informati on of wheat diseases by consulting experts and reference books, such as disease onset period, disease symptoms, growth morphology and corresponding images. Then, the neural network k nowledge of training simulation for BP network is obtained by preprocessing the disease information, and establishes the knowledge base.(3) Realize WEB accessible wheat disease diagnosis expert system. Through programming work to complete intelligent diagnosis and disease information browsing of 10 major wheat diseases, and the project is built on the TOMCAT server for user access and browsing.