Browse Prior Art Database

Defect Analysis of Steel Sheet Manufacturing Process by Datamining Methodology

IP.com Disclosure Number: IPCOM000015487D
Original Publication Date: 2002-Jan-01
Included in the Prior Art Database: 2003-Jun-20
Document File: 3 page(s) / 189K

Publishing Venue

IBM

Abstract

A procedure is disclosed that assists analysis and shooting of defects observed to "build" on the surface of steel sheet during the manufacturing process. The steel making process is known to be extremely difficult to analyze with deterministic methods, since it is a collection of highly nonlinear sub-processes. The proposed procedure circumvents this difficulty by using clustering algorithm that has the ability to "differentiate" data into strata according to the information entropy. The procedure assumes that the data describing the steel sheet manufacturing process are given as a collection of data items describing the alloy composition, the basic oxygen furnace (BOF) operation, the casting operation, the preheating operation, and the rolling operation. This is not an unrealistic assumption, as major steel making companies have long established extensive production control systems equipped with engineering data warehouses. See the figure 1 for the steel sheet manufacturing process. Continuous Casting Steel Sheet Manufacturing Process

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 57% of the total text.

Page 1 of 3

  Defect Analysis of Steel Sheet Manufacturing Process by Datamining Methodology

  A procedure is disclosed that assists analysis and shooting of defects observed to "build" on the surface of steel sheet during the manufacturing process. The steel making process is known to be extremely difficult to analyze with deterministic methods, since it is a collection of highly nonlinear sub-processes. The proposed procedure circumvents this difficulty by using clustering algorithm that has the ability to "differentiate" data into strata according to the information entropy.

The procedure assumes that the data describing the steel sheet manufacturing process are given as a collection of data items describing the alloy composition, the basic oxygen furnace (BOF) operation, the casting operation, the preheating operation, and the rolling operation. This is not an unrealistic assumption, as major steel making companies have long established extensive production control systems equipped with engineering data warehouses. See the figure 1 for the steel sheet manufacturing process.

Continuous Casting

Steel Sheet Manufacturing Process

Figure 1.

Defects on the surface of sheets are known to "form" during the rolling process. The proposed process applies clustering (independently) against:

1) The data items describing the alloy components and the steel making and casting

1

Pre-heating Furnaces

BOF

Slab

Scale Breaker Hot Rolling

Cold Rolling

[This page contains 1 picture or other non-te...