| Pests forecasting and control is the basis for the crop management to ensure the high yield and quality. With the development of information technology, computer science, such as data management, multimedia, GIS and network technology, have been widely used in the plant protection for monitoring, forecasting, decision support and diagnosis. Many crop database and data management system have been expored all over the world, but they were only applied for the specific pests in a certain region. There is a wide range of crops pests and diseases in China. In order to meet the needs of the informationization, internetization and standardization of pest's data, and the demands of daily work of the plant protection, a WebGIS-based pest warning & identify platform was built by the open source softwares (PHP, Apache, MapServer and MySQL). The main results were as follows.1. Establishment and implementation of an open network platform for the management of crop pests and diseasesA Web-based information management system of crop pests and diseases was built by the script language and database for the effective management and use of various pests' information. With an open design mode, seven modules were set up in this system, including dynamic tree-menu, online-design table, table management, scheduling reminder, data statistics and comparison, and graphics display. Therefore, users can customize the tables of database and build their own crop pest management system on this platform according to their demands. The system can manage different kinds of tables by the dynamic tree-menu, which has been connected through the module of online-design table. The functions of database management include data query, data sum, comparison and graphical display, and so on. 2. The web-based short-term early warning platform of pestsThe short-term forecasting plays a key role in the agriculture, because it can directly guide the farmers to control pests. It is a common method for the pest monitoring stations. This platform included the basic database of pest, such as the development duration at different temperatures, generation division and the occurrence degree. Users can add the actual data of local pests into this database. The authorized user can edit the records in this database.When the users put the investigation data into the pest database, the system will analyze the investigation data, the history data and the basic data, and automatically give some suggestions, such as the peak period of pest and the occurrence degree. With an open-style design mode, this system can provide a powerful platform for the users to create and mange their pests database. This system can process the originally data of field investigation and make a short-term pest forecast automatically. So the workload of pests forecast will be reduced. The prediction accuracy will also be improved significantly in this system.3. Development of pest's middle-and long-term forecasting models platform based on the data miningThe data mining method was used to build a middle-and long-term forecasting system for improving the status of enormous data and low-efficiency use of plant protection data. With the data produced by the system, users can easily select the factors and relative information from the web database, and then analyze the data with the statistics module integrated in the system, and ultimately build the middle-and long-term forecasting models. The system also provides the forecasting support for qualitative data through the decision tree method from ID3 algorithm. The technicians of plant protection can analyze the data without the knowledge of complicated database structure. The system is useful for the pest forecasting and control decision in different areas. The models can be modified and optimized as the records enlarged. The mathematical statistics modules of stepwise regression, fuzzy identification, and time-series analysis have been integrated in the system.4. Web-GIS based information display of pestThe pest database was linked with the spatial database through the Web-GIS, and the geographical map of the pest occurrence can be displayed in real time. The system can deal with the field investigation data automatically, and display the pest dynamic information by different colors or shape of symbols. The map can be published by internet automatically and be operated by zooming in or out. It is helpful for macro decision-making, analysis and early warning for crop pests and diseases.5. A Web-based platform for the diagnostic systems of crop symptomsInjured crop often show different symptoms and these can be used to identify the pests, disease, Phytotoxicity or fertilizer harm. These depend on the experience accumulated from production practice. They are just the expert experience, but have not been fully utilized for a long time. The knowledge representation of "IF-THEN" type was used in the expert knowledge base. The authorized user can put their experience or documents into the expert knowledge base through the internet. The documents include the pictures and descriptions of the pests. The retrieval type was used for inference engine. The three-tier distribution of web browser, web server and database were used in the system. With a concise interface, the users can easily build their own expert system without the complicated program knowledge and skills. The system unifies the users and builder of expert system and will have a wide application in the future.6. The establishment of the general taxonomic system for pestsThere are a large amount of pest species in plant protection, including millions of insects and thousands of weeds. According to the characters of species taxonomy, the dichotomous key of taxonomy was expressed as the records in the database. The tree-structure was used to manage the taxonomy information. The taxonomist can directly build and edit the tree structures through the internet. The taxonomic information can also be added into the database such as images, description and video of pests. The design of online editing makes the pictures and text discription of pest co-occurrence. The experts in different fields can be involved in the establishment of information, and the users can query and identify the species in this system. The taxonomy search mode with pictures and text is also suitable for the general technician of plant protection. This system is not only for the pests but for other species.In addition, the system includes the modules of import or export between excel files and database, the calculation of day degree, the variance analysis, the publication of pest information based on Google Map API and trajectory analysis. Import and export modules provided a web-based database interfaces for the user to store large amounts of historical data. The modules of calculation for day degree and variance analysis can process the experimental data. The module of Google Map API connects the map and the web database so that the pest's information can be displayed without the spatial data. The module of trajectory analysis simplifies the operation of work.Based on the thin client, this system has a low demand for the hardware and software. All the functions were implemented by the server, and all the softwares were the open source softwares Therefore, it is cheap to develop and maintain. With an open design mode, it is easy to operate and has a wide application. It is helpful to strengthen the information management of the crop pests through the network and to make scientific decisions. This research aimed at not only specific pests and provided the platform of data analysis and management for the technicans of palnt protection at three different levels (province, city and county). This system will also be improved in the future application.
|
PDF Full Text Request