Study On The Expert System Of Nutrition Diagnosis For Citrus Based On The Evidence Theory

The yield of citrus ranked at the first place in world fruits production. There weremore than 80 countries where citrus were grown in the world. As perennial tree fruit citruslongevity could be several decades and even more than a hundred years. To ensure highquality, bumper harvest and prolong fruiting years the element dosage and everycomponent balance must be taken into consideration to avoid plant malnutrition. It wasvery important for citrus production to give the plant diagnoses and cure exactly andtimely. Both the expert system and the D-S evidence theory used in citrus nutrition diagnosiswere discussed in this paper. Expert system for citrus nutrition diagnosis could be definedas a program to solve problem about citrus nutrition at expert's level, the primary characterof which was that its problem analysis and solution depended on the rule of expertexperience. It was very suited for such a field as agriculture where there were manycomplex and indeterminate issues. Based on the knowledge of citrus nutrition from experts, books and literatures etc. aintegrating knowledge base was established. The expert system for citrus nutritiondiagnosis was developed by using of expert system developing tool, the Xiongfeng 4.1software, taking "knowledge body and object module" as knowledge representation and"forward and backward" as mix reasoning method. In the practice of nutrition diagnosisthe primary result of diagnosis could be made in the system according to externalsymptoms of element deficiency in the citrus plant provided by users, then the furtherresult of diagnose could be made on the basis of the soil and leaf element analysis, and aproposal of cure measure cold be listed as a table. More than ten kinds of elements couldbe diagnosed in the system. After having been tested, the designed aim of the system had been achieved. Thesystem ran stably and reliably, the interface of person-computer alternation was friendly,and the visual operation was easy to study. The expert system for citrus nutrition diagnosiswould be available tools citrus production.