Material Consumption In Blast Furnace Trough Automatic Authentication System Design And Application Of Measurement Data

The material consumption of the blast furnace is a kind of important data which can guide the blast furnace production and calculate production target. The measured data of material consumption is of some characters, such as large quantity, multi-variety and strong correlation. Most process of measurement and authentication in our country is done manually, which can't keep up with the development of ERP systems. Combining the features and demands of the steel enterprise, this paper makes researches on the data identification and authentication methods, and designs an automatic authentication system for the material consumption data under the through of blast furnace.Based on the analyzing the interior characters of measurement data, considering the process mechanism and users' demands, this paper presents the general design scheme. To the material loss in large quantity and small quantity, an automatic authentication approach for abnormal data based on multi-fuzzy model is proposed. At the same time, an automatic identification method based on outlying data mining technology for abnormal data is proposed to the abnormal data resulted from exceptional coke data and ore data. At last, an intelligent model for analyzing and processing measurement data is build with the metal-material-balance method and the average method to make automatic amendment and repair to different abnormal data.An application software of automatic authentication system for the material consumption data under the through of blast furnace has been developed with Oracle 9i Database and PowerBuilder 8.0 developing tool. This new automatic authentication system includes functions of data-analyzing, automatic authentication, preferences, report forms and data-querying etc and has been put into used. By the operation in the scene of an enterprise ERP system, this system can increase the authentication efficiency, reduce the occurrence of the mistakes and has obvious application value. Simultaneity, its correctness and validity are been verified.