Design And Realization Of Sales Predicting And Blending System In The South Coal Mine

Compared with north of China, south coal mine has its own characters such as small scale, distributed location, complex-located coal bed, small output, great variety and different qualities. Because of these characters, the information management system used in north coal enterprises will become acclimatized when used in the south coal mine. As an essential link, combining its own characteristic with optimization of sales management is inevitable to meet market laws in the process of sales information technology.This paper’s study is based on Fujian Coal Electricity Co. Ltd. Firstly, the research background and status is introduced at home and abroad. Secondly, the author has got an overall design by analyzing the demand and feasibility for the system. Then, detailed design for each system module and database has been obtained on the basis. At last, the function of each module has been achieved by choosing the ASP.NET technology, Visual Studio2010application development platform and Microsoft SQL Server2005database.This paper is committed to solving problems about how to achieve product plan automatically and coal blending optimization. At first, the author has discussed the reason why GM(1, N) based on residual correction by back-propagation(BP) neural networks is chosen to research on sales prediction, and worked out the modeling steps with a real example. Next, the author has conducted research on single coal screening by combining greedy algorithm with Powell search method in coal blending optimization.Some modules of Sales prediction and deployment System in the south coal mine has been hybrid programmed by C#and matlab in this paper, which has shown the flexibility and diversity in designing thought. So far, the system has been used in Fujian Coal Electricity Co. Ltd successfully.