Design And Implementation Of Coal Enterprises Cost Management System Based On Data Mining

Cost management is a complex systems engineering. Presently, cost management for coal enterprises is mostly reflected in accounting, without advanced and intelligent management methods, and cost information is isolated, enterprises are unable to manage their costs in a well organized way. So, build an integrated cost management system for the enterprises, which combine with intelligent methods to transform the vast amounts of cost data into knowledge information, steer the cost management in the direction of multi-dimensional analysis sophisticated management is a necessary requirement for the enterprise development and market competition.In order to apply data mining on the cost management problem, this thesis studies on the basis of comprehensive study the cost management and data mining theory and technology. This thesis will do an in-depth analysis based on combined research status, as well as the practice in Fujian Coal and Electric Power Co., Ltd. Application systems, control and optimization ideological, and a number of key points of cost management were studied by integration of multidisciplinary knowledge。Firstly, this paper built the multidimensional and multilevel cost management model based on data mining and integrated the information between the cost management and ERP/MES system. Through the ETL process to ensure the reliability and accuracy of the cost data, then explains how to implement the enterprise cost management and completion of specific management functions under the model. Secondly, the cost management system was designed and implemented with the required enterprise cost management situation and needs, which promoted the enterprise automation, intelligence and information technology processes. Finally, the paper focuses on the research and realization of the data mining module. This paper uses the decision tree algorithm to mine the difference of the data to explore the factors that affect the cost, and construct mathematical model to schedule production resource by using the quality, efficiency and cost as threshold values. A multi-objective relative degree of optimization model for quality-efficiency-cost was built. Solving the optimization model by Non-dominated Sorting Generic Algorithm (NSGA-Ⅱ), the optimization results greatly eliminate coupling between different indicators and improve the scientific and flexible of the production resource planning and production group selecting. This method has practical significance to improve the enterprise decision-making and scheduling production resource.