Research On Tea Plant Diseases And Insect Pests Expert System Based On CBR-RBR Integration And Deep Learning Method

With the implementation of the strategy of "one belt and one road",tea production and export volume ranked first in the world.Tea,as an important commodity in the traditional "one belt and one road",has attracted the attention and attention from all walks of life.Therefore,the diagnosis and prevention of tea plant diseases and insect pests has become a hot issue in universities and research institutes.With the deep integration of artificial intelligence and agriculture,it is of great significance to explore the application of CBR-RBR(case-based and rule-based reasoning)integration and deep learning methods in the diagnosis of tea plant diseases and insect pests,so as to improve the quality and yield of tea in China.The main research contents are as follows:1.Integrated reasoning method based on CBR-RBR(case-based and rule-based reasoning method).Previous reasoning methods of expert system for tea plant diseases and insect pests often adopt rule-based reasoning method.A few people use case-based reasoning method,while the research methods of combining CBR and RBR for reasoning are less studied.In this study,RBR reasoning method,CBR reasoning method and CBR-RBR integrated reasoning method were applied to the diagnosis of tea plant diseases and insect pests.The results show that CBR-RBR integrated reasoning strategy can effectively improve the accuracy of system diagnosis compared with RBR reasoning alone and CBR reasoning alone.The average accuracy of CBR-RBR integrated reasoning is 83.3%.2.Diagnosis of pests and diseases based on image recognition.Expert systems based on text retrieval often suffer from illnesses that are unclear,inaccurate or difficult to describe.Therefore,the diagnosis based on text information retrieval alone can not meet the actual needs.By adding image recognition technology to the traditional expert system,farmers can be assisted to diagnose more accurately.In this study,the convolution neural network is used to preprocess the acquired images of tea plant diseases and insect pests,extract automatic features,and achieve the final image recognition and classification.According to the test results,the average recognition time of this system is 1.169 s,which accords with the perception of ordinary people.The Average hit rate of image recognition is 95%.The results show that this method can better realize the diagnosis and classification of pests and diseases.According to the research results,an expert system for diagnosis of tea plant diseases and insect pests is developed.The system is based on B/S architecture and developed with ASP.NET technology.The system includes the modules of self-help diagnosis of tea plant diseases,browsing of tea plant diseases,browsing of tea plant pests,expert diagnosis of tea plant diseases and insect pests,and recognition of tea plant diseases and insect pests based on image processing.Through preliminary application,it can meet the requirements of diagnosis and identification of tea plant diseases and insect pests in tea production.