Research And Application On The Development Of Agricultural Monitoring And Early Warning Database System

Those have brought the reality of demand for China’s agricultural monitoring and early warning that the change of natural, economic and social environment, as well as unexpected events. Agricultural monitoring and early warning was the work that tracking and monitoring the related fields of agricultural production, circulation and consumption, analyzing, judging and forecasting the future trends by advanced models and efficient data, providing early warning for the situation which caused by the linkages influences. Therefore, it has important practical significance and social value to research and develop the agricultural monitoring and early warning database application system, to publish monitoring and early warning information timely for market participants, to provide support for agricultural monitoring and early warning practices and decision departments, to improve the level of agricultural information service, to serve the research of monitoring and early warning methods, state macro-control and policy-making.Through using overseas and domestic research status about agricultural monitoring and early warning system, key technologies, agricultural data sources, standardization and preprocessing, distributed database systems and agricultural application systems for reference, this study learned some advanced institutions and experiences and compared different model theories, data collection technologies, industry standards, data preprocessing technologies, database system types and the related application. The main task of the study was to carry out research on agricultural monitoring and early warning database system, then design and build the system and develop application analysis for it.First of all, this study focused on considering the needs of business logic and model technology of agricultural monitoring and early warning, and analyzed the data comprehensively. The data were divided in four types:dynamic monitoring data, statistical data, invoked data and configured data. The detailed analysis for the data sources included the methods of collection, integration framework of agricultural Deep Web data, data characteristics and problems. In order to solve data quality problems exist, the paper studied how to apply the relevant agricultural data standards and how to preprocess the source data and the data based on business requirements.Secondly, after the analysis of data and system requirements, the study designed a framework for agricultural monitoring and early warning database system. It analyzed each of databases according to distributed structure, achieved conceptual and logical design of model configuration database and proposed a multidimensional data model of OLAP data warehouse. Some key ingredients were designed and implemented by program such as index lookups in basic data warehouse, storing and getting slice data by OLAP, model management in model management database and data version management.Finally, the study designed agricultural monitoring and early warning database system, which was a preliminary system including related functional modules as Agricultural Monitoring and Early Warning Data Center, Agricultural Market Price Analysis System and Agricultural Monitoring and Early Warning Model Management System. It also raised the service items and application modes. What’s more, taking the case of sugarcane, It conducted an empirical study on how to receive the result of yield forecasting by the application system.