Decision Support System For Citrus Production Management

China is one of the origin of citrus and has abundant citrus resources.In recent years,citrus farming has developed rapidly,and has become a pillar industry for China’s export of foreign exchange and poverty alleviation in mountainous areas.However,citrus is mostly grown in the south,and the warm and humid climate in these areas has led to the frequent occurrence and spread of citrus pests and diseases,which have seriously jeopardised citrus production activities.Due to this problem,many farmers have adopted the so-called "insurance control" strategy,increasing the number and concentration of chemical pesticides used,resulting in excessive use of pesticides,deterioration of citrus quality and environmental degradation.It is crucial for frontline citrus farmers to identify pests and diseases in the orchard and to apply pesticides at the right time to manage and control them.With the rapid development of information technology,it is reasonable to apply "flying paddle" image recognition technology and LBS technology to the field of agricultural information technology to manage and control citrus orchards and promote the process of agricultural modernisation.At the same time,the development of WeChat applets has brought users a convenient and friendly user experience.Combining WeChat applet technology with agriculture can break the "last mile" bottleneck of agricultural information technology for agriculture and rural areas,and can accelerate the development and application process of agricultural information technology.In this paper,through the WeChat developer platform,an auxiliary decision support system for citrus production management based on WeChat applets is developed,realising three main functions: citrus business information display,citrus pest and disease image recognition and citrus geographic information management.The main studies are as follows.(1)Firstly,the citrus business information display function provides users with citrus information,citrus disease information and citrus pest information,and users can view the content of specific information details by clicking on the information in the information list.The information is collected through data preparation,stored in MongoDB database,deployed on Ali cloud server,called through RESTful API generated by Swagger design,and displayed on the page with the help of wx:for component to achieve the information display function.(2)Secoundly,the citrus pest and disease image recognition function provides image recognition of pests and diseases found by users,who can quickly access pest and disease knowledge and control measures by using this function in their actual production activities.The pest and disease sample images are collected through data preparation,and the pest and disease model is trained with the AutoDL Transfer model through the "flying paddle" framework in the EasyDL platform.(3)Lastly,the citrus geographic information management function provides users with a basis for citrus fertility grading.Users can obtain the fertility grading of the main citrus production areas,and make appropriate fertiliser type selection based on the location grading obtained.The data is prepared by collecting and testing the sampling points in the main citrus production areas,storing the sampling point information based on PostgreSQL and PostGIS databases,formulating the citrus soil nutrient grading standards,interpolating the fertility grading situation in the main citrus production areas with the kriging interpolation algorithm,and using LBS technology to display the soil fertility grading map and fertiliser sampling point data in the main citrus production areas nationwide to realise the citrus Geographic information management function.The citrus production management support system is based on WeChat applets and has the advantage of being convenient and efficient,helping users to obtain citrus business information,carry out pest and disease image identification and obtain citrus fertility grading level information,providing management decision support functions for citrus production activities,reducing citrus pest and disease hazards,providing fertiliser recommendations,improving citrus quality,reducing environmental pollution and increasing farmers’ income.Provide powerful technical support for citrus production activities.