Tadaaki Iwamura

Blast Furnace Control by Artificial Intelligence

Abstract Kawasaki Steel Corporation has recently developed a new blast furnace control system adopting an expert system, a kind of artificial intelligence technology. Knowledge representation in this system is designed to reflect knowledge structure of experts with fidelity, and is written in production rules. It is practiced in an engineering work station for exclusive use of AI, and gives operation guidance to operators every five minutes. Evaluation of this system was made by comparing operators and this system in terms of judgment in actual operation, and good results were achieved.

BLAST FURNACE CONTROL BY ARTIFICIAL INTELLIGENCE

Kawasaki Steel Corporation has recently developed a new blast furnace control system adopting an expert system, a kind of artificial intelligence technology. Knowledge representation in this system is designed to reflect knowledge structure of experts with fidelity, and is written in production rules. It is practiced in an engineering work station for exclusive use of AI, and gives operation guidance to operators every five minutes. Evaluation of this system was made by comparing operators and this system in terms of judgment in actual operation, and good results were achieved.

The New Instrumentation System at Chiba No. 6 Blast Furnace and its Contribution to Operation

Abstract A distributed and hierachical computer system was developed for the control of Chiba No. 6 Blast Furnace which is one of the largest and latest blast furnaces in the world. Fully utilizing the digital instrumentation and communication control techniques, the system is capable of handling a vast amount of process data and realizing some complicated process control. The system contributes to the establishment of high standard of the blast furnace operation. Some examples of the way the system is contributing to operation include: the instrumentation and control of gas flow distribution inside the blast furnace; quantification and optimization of pattern information by means of discriminant analysis; overall control of furnace control by the Go-Stop system; hot stove combustion control; level control of raw material bins, and torpedo car control system.

Software Quality and Productivity Improvement Techniques for Process Computer Systems at Mizushima Works, Kawasaki Steel Corporation

Abstract Process computers have a long history of use in the steel industry, where they have made a remarkable contribution to product quality and production efficiency. Moreover, process computer functions have become increasingly sophisticated and complex in response to the requirements of high-level line automation, making the improvement of software quality and productivity an essential task. Since 1987, Kawasaki Steel has actively promoted the development of techniques for the improvement of software productivity. The main software techniques developed are as follows. (1) Autonomous distributed systems An autonomous distributed system was put into practical application in the steel industry for the first time at Mizushima Works, resulting in improved software portability, which had been inadequate with process computer systems in the past. (2) Reuse of software Better efficiency was achieved in software design and production by the use of parts management practices developed for software specifications and programs on the basis of improved software portability. (3) Improved development environment A network system is essential as infrastructure for the reuse of software; a high-speed/highcapacity FDDI was introduced for this purpose. A multi-vendor development environment was constructed by linking process computer systems, workstations, and personal computers. These techniques were applied to a CGL system and to two continuous annealing line systems. The results include a reduction in software bugs and a two-fold increase in productivity. This paper presents an outline of software techniques developed by Kawasaki Steel Corporation.