The Research And Achievement Of Regional Energy Consumption Ews Based On Data Warehouse

With the fast development of economy, the problem of energy development has become a national strategy. Energy management has being a significant research issue. At present, about the energy-saving and cost-reducing research of enterprise is relatively more, but the energy research whose research object is some entirety or some region is less, at the same time, for the energy consumption macroscopic management of government and the energy-saving of enterprise, the analysis and early warning of regional energy are significant. Current energy management more and more forwardly expects to discover the regular knowledge of region, and proposes stricter request to the intelligent analysis and decision of the data of energy consumption.Basing on the energy management of region and enterprise, in allusion to the problems of the regional energy management that about the data analysis and early consumption. This thesis adopts the technology of data warehouse and the algorithm of data mining which results in favorable practical effect The main work and achievements are as follows:1. In allusion to the multi-dimensionality of regional energy consumption data, this thesis brings in the technology of data warehouse, confirms the theme of energy consumption early warning by means of conceptual model design, selects star model as the model of data warehouse by means of logical model design, and by means of physical model design, confirms data base’s index, storage, structure and so on. Consequently, building the data warehouse of regional energy consumption, and doing OLAP multi-dimensional analysis on the data base.2. In allusion to the fault existed in traditional early warning, this thesis proposes the correlational early warning of energy consumption data with the association rules and by means of studying the algorithm of association rules, combining the multi-dimensionality of data, proposes one kind of layered Apriori algorithm which has space-time characteristic, this algorithm improves the mining speed of classic Apriori algorithm, which is suitable for the mining of relative early warning knowledge.3. In allusion to the utilization of energy efficiency, this thesis considers to research the multi-dimensional data under some indexes, adopts PSO and K-means mixed clustering algorithm that is suitable for the regional energy consumption EWS to do the unsupervised clustering, doing early warn for the far away item from the clustering center, thus benefits management to adopt effective measure, finally improving the use ratio of regional energy.4. Based on the enterprise’s survey, this thesis researches and comes true the energy consumption EWS, by means of the design of system function, flow path and data structure, constructing regional energy consumption EWS and enterprise energy management system; In the course of system implementation, based on the data warehouse, leading the improved Apriori algorithm to do the correlational early warning, leading PSO-K mixed clustering algorithm to do clustering early warning, the generated knowledge can provide decision support for regional management.