Study On A MCS-based Expert System Of Nutrition Diagnosis For Crops

This expert system is different from the tradition design mode of the expert system because it is run on TYZ-3 intelligent soil nutrition gauge which is controlled by the single chip microcomputer other than on desk-top computer. The program of this nutrients diagnostic system was designed and developed by both language of compile and MCS- C51 combined with the applying knowledge of MCS and expert system. We founded a MCS-controlled system which can simulate the diagnosis progress of expert for crops on the base of collecting and analyzing the symptoms of deficient elements on several common crops, such as apple, wheat, corn, tomato and cucumber.This paper briefly summarized the development of the agriculture expert system inside and outside of country and its application in nutrient diagnostic. The several key points were highlighted in this paper: the establishing of the scheme; knowledge change, knowledge express, knowledge storage and knowledge search; and further more optimizing the process of reasoning by using power-added uncertainty reasoning design. The interface of human and machine along with the situation of system running also has been introduced.In order to improve the system efficiency of diagnose, we tried to decompose the knowledge into the smallest cell and keep its logical structure clearly at the same time. By using express method of produce modal rule which has being used by artificial intelligence expert system to express the system's knowledge, using the method of power-added uncertainty reasoning to complete the process of nutrient diagnosis, we successfully get the proper treat scheme matched with the symptoms in the end.