인공신경회로망을 이용한 고로조업지원 전문가 시스템 연구

최태화; 윤인섭

HWAHAK KONGHAK, Vol.29, No.3, 270-283, June, 1991

Export Citation

인공신경회로망을 이용한 고로조업지원 전문가 시스템 연구

An Expert System to Aid Blast Furnace Operation Using Artificial Neural Network Techniques

최태화, 윤인섭

초록

철을 생산하는 일관제철소의 고로공정을 대상으로 조업지원을 위한 전문가 시스템을 개발하였다. 본 연구에서는 종래의 규칙기반 전문가 시스템이 가지고 있는 자체 학습능력의 미흡이나 지식표현 등의 문제점을 극복하고자 인공신경로망을 응용한 노황이상진단방법을 제시하였다. 진단의 선명성 향상과 여러 가지 이상상태의 진단에 대처하는 진단의 강건성을 높이기 위해 고로에 설치된 각종 sensor data를 전처리하여 사용하였다. 이상진단을 위해 입력층, 중간층, 출력층이 있는 3층으로 된 back-propagation type의 인공신경회로망을 구성하였다. 진단의 구조는 우선 노황의 이상유무를 판정하고 이상이 있으면 그 종류에 따라 구체적인 진단을 하는 계층구조로 구성하여 신속하고 정확한 진단을 할 수 있게 하였다. 진단결과인 노황이상의 종류와 진행정도에 따라 조업실적 분석결과를 근거로 마련한 action guidance를 CRT화면을 통해 제시하였다. 과거 이상현상이 발생했던 전후의 조업 data를 사용한 본 시스템에 의한 진단결과와 그 당시의 조업경위를 비교해 보았다. 노황의 점진적인 변화로 인해 조업자는 초기의 이상상태를 인식하지 못하고 아무런 조치를 취하지 않은 반면에 본 전문가 시스템은 조기에 이상상태를 잘 파악하여 적절한 action을 제시하였다.

An expert system has been developed to support blast furnace operations in the integrated iron and steelmaking works. In the present study, a technique to diagnose furnace abnormality, using the artificial neural network, is proposed to overcome problems of the conventional rule-based expert system, such as lack of in-system automatic regulation and the limits of knowledge expression, etc. In order to enhance diagnostic resolution and robustness to counteract various furnace abnormalities, the data acquired by several sensors installed in the blast furnace were pretreated before using them as imput sources for diagnosis. For the diagnosis of abnormal conditions, back-propagation type of artificial neural networks were constructed, which consist of input, hidden and output layers. Diagnostic procedures are structured hierarchically, that is, first the diagnostic network finds the abnormal conditions and then carries out a detailed diagnosis ac-cording to the type of abnormality. In this study, an action guidance is suggested through the analysis of actual operation results according to the type and progress of the abnormal conditions. The expert system was tested using the operation data obtained from the abnormalities in a real blast furnace. Without the expert system, the operator could not recognize the initial irregularity due to the gradual change of furnace conditions. On the other hand the expert system could find the initial furnace abnormality and suggested appropriate counteractions.

[References]

오세영, 전기학회지, 38(2), 17 (1989)
일본화학공학회편: "화학공학에 있어서 Expert System," 화학공업사, 13 (1988)
포항제철(주): "기술표준서 (고로공정)," (1987)
Aharonian G, J. Neural Network Comput., 53 (1989)
Dietz WE, Kiech EL, Ali M, J. Neural Network Comput., 5 (1989)
Dogherty ER, Giardina CR, "Mathematical Methods for Artificial Intelligence and Autonomous Systems," Prentice-Hall, London, 59 (1988)
Eberlein S, J. Neural Network Comput., 5 (1989)
Finch FE, Kramer MA, 10th World Congress on Automatic Control, 6, 298 (1987)
Fuji-Facom Control Co.: "EIXAX Manual," Fuji-Facom Co., 5 (1989)
Himmelbrau DM, "Fault Detection and Diagnosis in Chemical and Petrochemical Process," Elsevier (1978)
Hirata T, Yamamura K, Morimoto S, Takada H, 전기학회 산업응용부문 전국대회 발표집(일본), 807 (1989)
Hoskins JC, Himmelblau DM, Comput. Chem. Eng., 12(9), 881 (1988)
LISPE: "Knowledge-based Expert Systems in Process Engineering," Dep. of Chem. Eng., MIT, U.S.A., Summer (1990)
McClelland DE, Rumelhart and PDP Research Group: "Parallel Distributed Processing," 6th, MIT Press, Cambridge, Ma. (1987)
Oh JK, Ph.D. Dissertation, Seoul National Univ., Seoul, Korea (1990)
Pao YH, "Adaptive Pattern REcognition and Neural Networks," Addison-Wesly Publishing Co., 113 (1989)
VerDuin WH, J. Neural Network Comput., 46 (1990)
Venkatasubramanian V, Chan K, AIChE J., 35(12), 1993 (1989)
Watanabe K, Matsuura I, Abe M, Kubota M, Himmelblau DM, AIChE J., 35(11), 1803 (1989)
Yagi JI, CAMP-ISIJ, 2, 2 (1989)

[Referenced By]

[1] 오세영, 전기학회지, 38(2), 17 (1989)

[2] 일본화학공학회편: "화학공학에 있어서 Expert System," 화학공업사, 13 (1988)

[3] 포항제철(주): "기술표준서 (고로공정)," (1987)

[4] Aharonian G, J. Neural Network Comput., 53 (1989)

[5] Dietz WE, Kiech EL, Ali M, J. Neural Network Comput., 5 (1989)

[6] Dogherty ER, Giardina CR, "Mathematical Methods for Artificial Intelligence and Autonomous Systems," Prentice-Hall, London, 59 (1988)

[7] Eberlein S, J. Neural Network Comput., 5 (1989)

[8] Finch FE, Kramer MA, 10th World Congress on Automatic Control, 6, 298 (1987)

[9] Fuji-Facom Control Co.: "EIXAX Manual," Fuji-Facom Co., 5 (1989)

[10] Himmelbrau DM, "Fault Detection and Diagnosis in Chemical and Petrochemical Process," Elsevier (1978)

[11] Hirata T, Yamamura K, Morimoto S, Takada H, 전기학회 산업응용부문 전국대회 발표집(일본), 807 (1989)

[12] Hoskins JC, Himmelblau DM, Comput. Chem. Eng., 12(9), 881 (1988)

[13] LISPE: "Knowledge-based Expert Systems in Process Engineering," Dep. of Chem. Eng., MIT, U.S.A., Summer (1990)

[14] McClelland DE, Rumelhart and PDP Research Group: "Parallel Distributed Processing," 6th, MIT Press, Cambridge, Ma. (1987)

[15] Oh JK, Ph.D. Dissertation, Seoul National Univ., Seoul, Korea (1990)

[16] Pao YH, "Adaptive Pattern REcognition and Neural Networks," Addison-Wesly Publishing Co., 113 (1989)

[17] VerDuin WH, J. Neural Network Comput., 46 (1990)

[18] Venkatasubramanian V, Chan K, AIChE J., 35(12), 1993 (1989)

[19] Watanabe K, Matsuura I, Abe M, Kubota M, Himmelblau DM, AIChE J., 35(11), 1803 (1989)

[20] Yagi JI, CAMP-ISIJ, 2, 2 (1989)