Research On Data Warehouse Construction And Visualization Technology Based On Cotton Storage

China National Cotton Reserves Corporation has more than a dozen of direct warehouses which are under the State’s investment, and there are nearly two hundred substitution warehouses. The cotton warehouses are sited across the country mainly in the major cotton production and marketing areas and the storage capacity reaches about1.3million tons. Such a large cotton warehouse management information system will produce large amounts of warehousing environment data and the data plays an important role in a short time during the company’s decision-making center takes real-time regulation. But the data over the past period of time no longer has real utility. The topic is to build a cotton data warehouse by using data extraction, transformation, data loading and other model treatments. Mining the large amount of historical data in the cotton data warehouse then draw implicit, previously unknown but potentially valuable information. The information can be used to assist cotton warehouse administrators to improve cotton warehouse environment, or help corporate decision makers to adjust the cotton flow condition to make appropriate decisions.The article is based on the project "The key technologies of cotton storage feature detection safety and quality management and application demonstration". The article’s research focus on building the cotton data warehouse and multidimensional visualization analysis, as well as cotton warehouse warning display based on data mining. The concrete contents are as below:1. Build cotton data warehouse. Design the data warehouse framework basing on "The key technologies of cotton storage feature detection safety and quality management and application demonstration". Then clean, transform, load the cotton historical data.2. Display the cotton data warehouse cube. Design different levels of data abstraction according to different users’query. Organize data in multidimensional way and display the data in a multidimensional way. We mainly use two-dimensional tables, data distribution chart, curve or any combination of the methods to provide decision support intuitively and quickly for decision-makers.3. Analysis and display the cotton warehouse environment data by using ArcGIS software. Using BP network algorithm to analysis cotton warehousing environmental data and cotton level data. Then get three levels of environmental impact on cotton level. At last give tips for warehouses which have risk.