An Agricultural Insect Pests Expert System Reasoning Machine Based On Case And Fuzzy Technology

The agricultural insect pest is one of the most important limiting factors to the agricultural production. Also it is the primary cause which makes our agricultural economy suffering the heavy loss. However, almost every agricultural producer is unclear to the insect pest, blind preventing and controlling, leads to the excessive waste of the agricultural chemicals, and brings the serious burden and the pollution to the environment. Furthermore, our country is a large agricultural nation, the talented person who masters the agricultural technology is very scarce, the dissemination and application cycle of the advanced science and technology is long, and the develop speed is slow, the agricultural production and the management level are low too, all these have restricted the agricultural production and the sustainable development seriously. To make a reasonable diagnosis plan of the insect pest is a comprehensive complex process, which includes the insect pest characteristic analysis, the characteristic extraction, the reasoning algorithm research as well as accurately gives the conclusion and a preventing and controlling measure, its primary mission is the insect pest qualitative recognition. The complete quality and efficiency of this process decides the receive level of the agricultural producer's production, also has manifested the practical significance of the plant protection. In view of the realistic situation, this research unifies the present research situations of the domestic and foreign, based on the insect taxonomy as well as the plant protection knowledge, recording to the actual agriculture insect pest diagnosis demand of our country, takes solving the difficult production problem as the goal, uses the case and fuzzy reasoning mechanism, carries on the research of the reasoning machine of the agricultural insect pest diagnoses expert system, has the very strong guiding sense to the reality.In order to fulfill the actual request of the insect pest characteristic diagnosis to the expert system, considering the realistic characteristics and requirements of the insect pest characteristic diagnosis, a hybrid reasoning method integrating case-based reasoning with the fuzzy technology is presented for the first time, and how the reasoning machine works is given. The method is introduced to the design of the insect pest diagnose expert system. The description of the knowledge expresses of the case-based and fuzzy-based mechanism is given. A reasoning mechanism integrating case-based reasoning with the fuzzy-based reasoning is proposed and the integration mode is illustrated in detail. The thesis mainly completed the insect pest qualitative recognition, the research of the knowledge expression of the insect pest diagnosis expert system, the design of the system reasoning machine as well as the design of the explanation function module.This research uses the object-oriented advanced language Delphi as the development language, which is good at the network and database operation, uses the necessary Desktop tool to design the DB database, uses the newest pattern recognition technology to develop the reasoning machine of the case and fuzzy based expert system which have the self-renewal ability, the case self-study ability, the case similarity judgment ability, the case revision ability as well as robustness ability, has made a beneficial attempt in the overlapping domain of plant protection specialized knowledge and the computer technology, has mainly yielded the following result:The thesis mainly carries on the research of the reasoning machine, which integrates the case-based reasoning with the fuzzy technology , uses the"case-reasoning first fuzzy-reasoning second"reasoning mechanism to simulate the human thought to complete the inference process, and more accurate solution is given for interdisciplinary complex realistic questions. The system attempts to use case-fuzzy concept to express the domain knowledge, and carries on the reasonable numerical to the characteristic parameter to guarantee the accuracy of the reasoning, gives the user the final conclusion by the fuzzy language and the confidence level form, and give the system diagnosis result to assists the user .The system uses the modulation structural design, which makes the system maintenance more easily, just need the expansion to the specific module, has provided the connection for the re-development.The practice proved that, the case and the fuzzy reasoning thinking method and the expert system based on the special reasoning mechanism has the good inference ability and the extremely high processing efficiency, can aim at the concrete insect pest characteristic, effectively carries on the recognition and the diagnosis to the plant insect pest, provides the powerful safeguard for the agricultural production.