The Research Of Data Warehouse Platform For Line Losses Based On Oracle

With the line losses management system in power industry continued to run, power companies have accumulated a large amount of historical data of line losses. The traditional information management methods have been unable to scientifically and effectively deal with and use of such information, but also should not have to identify potential data information with economic value. Data warehouse theory can do the correlation analysis for a great variety of business database and optimize existing resources, integration of existing systems of information when applied in power line losses analysis, exert overall advantages for different levels of management personnel to provide effective decision support.This paper considers the application in practice. It designed and implemented the data warehouse platform of line losses by using of the data warehouse theory and cleaned the duplicated records in data warehouse with the cleaning strategy.Firstly, in the light of the structure and characteristics of the popular line losses calculation and analysis systems, this paper designs the architecture for line losses data warehouse platform by DB-ODS-DW (Database-Operational Data Store-Data Warehouse) three-tier model and gives the detailed modeling process for line losses analysis data warehouse platform by star-shaped model. It is difficult to achieve real-time analysis and mining applications through the traditional two-tier architecture. The low efficiency and supporting ability of real-time and difficult issues such as data integration etc problems of DB-DW (Database-Data Warehouse) two-tier structure was well solved by introduction the ODS(Oracle Warehouse Builder) to architecture.Secondly, by using of the Oracle database and OWB (Oracle Warehouse Builder) instrument to set up the data warehouse with the target of line losses analysis in this paper. It makes the information integration of the business information in power industry and realizes the multidimensional analysis of business data, and provides effective decision support for managers.Finally, this paper gives some advices for improving the problems in the"scheduling,detecting,merging"algorithms of duplicate elimination. The improved duplicated records elimination algorithm has effectively promoted the efficiency of scheduling record on the environment that record matching rate was keeping high. In detecting duplicated records, it takes into account 4 factors. For instance, the number of characters, the existing frequency of character be found, the importance (weight) of field in records ,the Chinese semantic and the semantic focus is always in the back location etc. It makes the duplicated records detecting algorithm more accurate and healthy. In merging duplicated records, this paper uses both the cluster algorithm and practical algorithm which has greatly improved the speed of duplicated records cleaning and has considerably reduced the workload of users.