Development Of Gas Station Integrated Analysis System Based On Data Warehouse

A gas station is a site that specially provides refueling and other value-added services for cars and other motor vehicles. Gas stations are the basic component of the Sales Department of petro-chemical enterprise. The vast majority of sales business data comes from the gas stations.The information management system of the petrochemical enterprise holds a great many basic data, oil data, retail data, leasing data and other information. However, due to the huge amount of data and complex data structure, the information department cannot directly use these data to provide support for enterprise leaders. Therefore, it is necessary to establish an information management system based on data warehouse to manage the data.In the design of the integrated analysis system of gas station, the theory of data warehouse is the foundation in the whole system. The improvement and perfection of the traditional data warehouse structure is the most important work of the construction of the sales enterprise data warehouse. Combined with the characteristics of petrochemical industry, BW on HANA, which is the data warehouse architecture based on the high performance data analysis equipment, is integrated the approach to the design of this architecture to the implementation of the project, which is the main purpose of the thesis. In order to solve the deficiency of the traditional database management system, the paper focuses on the improvement of the data increment mechanism and the optimization of the gradient dimension rule. In improvement of the increment mechanism, the paper focus on the two methods of incremental queue and tag field, and puts forward the self-updating method improved increment mechanism. In optimization of the gradient dimension, the paper focuses on three modes of reference of current time, reference to the intersection of past and the current and reference to the union of past and current. Through the use of virtual feature, the gradient dimension rules were optimized. In the oil product sales forecast module, the moving average algorithm and the triple exponential smoothing algorithm are mainly studied, and the calculation results of the two algorithms are evaluated. Finally, the exponential smoothing method is improved by optimizing the selection of parameters, which can achieve better prediction results.Through the development of the data warehouse based integrated analysis system of gas station, the basic process of data warehouse development is understood and mastered. According to a large number of business history data, the enterprise’s leaders can get a full range of query services. With the specific legible data support, enterprises can set out plans of purchase, sell and inventory more scientifically and rationally. As a result, greater economic benefits will be created.