Doctor of Philosophy
School of Computing and Information Technology
Steel is a fundamental raw material for all industries. It can be widely used in vari-ous fields, including construction, bridges, ships, containers, medical devices and cars. However, the production process of iron and steel is very perplexing, which consists of four processes: ironmaking, steelmaking, continuous casting and rolling. It is also extremely complicated to control the quality of steel during the full manufacturing pro-cess. Therefore, the quality control of steel is considered as a huge challenge for the whole steel industry. This thesis studies the quality control, taking the case of Nanjing Iron and Steel Group, and then provides new approaches for quality analysis, manage-ment and control of the industry.
At present, Nanjing Iron and Steel Group has established a quality management and control system, which oversees many systems involved in the steel manufacturing. It poses a high statistical requirement for business professionals, resulting in a limited use of the system. A lot of data of quality has been collected in each system. At present, all systems mainly pay attention to the processing and analysis of the data after the manufacturing process, and the quality problems of the products are mainly tested by sampling-experimental method. This method cannot detect product quality or predict in advance the hidden quality issues in a timely manner. In the quality control system, the responsibilities and functions of different information systems involved are intricate. Each information system is merely responsible for storing the data of its corresponding functions. Hence, the data in each information system is relatively isolated, forming a data island. The iron and steel production process belongs to the process industry. The data in multiple information systems can be combined to analyze and predict the quality of products in depth and provide an early warning alert. Therefore, it is necessary to introduce new product quality control methods in the steel industry. With the waves of industry 4.0 and intelligent manufacturing, intelligent technology has also been in-troduced in the field of quality control to improve the competitiveness of the iron and steel enterprises in the industry. Applying intelligent technology can generate accurate quality analysis and optimal prediction results based on the data distributed in the fac-tory and determine the online adjustment of the production process. This not only gives rise to the product quality control, but is also beneficial to in the reduction of product costs. Inspired from this, this paper provide in-depth discussion in three chapters: (1) For scrap steel to be used as raw material, how to use artificial intelligence algorithms to evaluate its quality grade is studied in chapter 3; (2) the probability that the longi-tudinal crack occurs on the surface of continuous casting slab is studied in chapter 4;(3) The prediction of mechanical properties of finished steel plate in chapter 5. All these 3 chapters will serve as the technical support of quality control in iron and steel production.
Li, Fucun, Towards A Computational Intelligence Framework in Steel Product Quality and Cost Control, Doctor of Philosophy thesis, School of Computing and Information Technology, University of Wollongong, 2021. https://ro.uow.edu.au/theses1/1299
FoR codes (2008)
0806 INFORMATION SYSTEMS, 0801 ARTIFICIAL INTELLIGENCE AND IMAGE PROCESSING, 0807 LIBRARY AND INFORMATION STUDIES, 0905 CIVIL ENGINEERING
Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.