Font Size: a A A

Research On Airport Energy Information Management Based On Data Warehouse

Posted on:2018-07-22Degree:MasterType:Thesis
Country:ChinaCandidate:C LiuFull Text:PDF
GTID:2322330533460087Subject:Control Science and Engineering
Abstract/Summary:PDF Full Text Request
The airport field has accumulated a lot of energy data in its daily operations.With the continuous development of the airport scale and the rapid increase in airport energy consumption,the science management of airport energy information become more and more important,and the traditional analysis which based on database statistical query can't meet the demand of airport energy information management and analysis.Data warehouse technology is applied to the airport energy information management,the airport energy information data warehouse is constructed,and the clustering mining is carried out on the basis of the data warehouse.Firstly,the demand of the airport energy management data warehouse system was analyzed and the data warehouse model was designed.Two main themes of the airport energy management data warehouse was built,and further the logical model structure and the physical model structure for the two topics was established.According to the characteristics of airport energy data,the two-steps ETL structure was designed for the airport energy data processing.Achieved remote heterogeneous database access and data extraction and loading of data conversion and data cleansing,data integration.Secondly,Clustering analysis based on the airport energy data warehouse to realize the effective information mining of historical data was presented,and the K-Means algorithm was improved in the K value selection and initialization of the clustering center.The multidimensional energy consumption time series was constructed with the monthly energy consumption value as the data point,and it is used to determine the energy consumption mode of the energy unit.Finally,based on the improved K-means algorithm,the airport energy data w analyzed by cluster experiment.The validity of the improved K-means algorithm is tested by the classical data set.The monthly energy consumption data of the airport was clustered to establish the energy consumption benchmark value for different groups of units.Finally,the group segmentation mode was established by energy consumption level and energy consumption rate Sequence clustering.
Keywords/Search Tags:Data Warehouse, ETL, K-Means Clustering, Time Series, Group Segmentation
PDF Full Text Request
Related items