基于集成案例推理方法的RH精煉鋼水終點溫度預測

馮凱; 徐安軍; 賀東風; 汪紅兵

doi:10.13374/j.issn2095-9389.2018.s1.023

基于集成案例推理方法的RH精煉鋼水終點溫度預測

doi: 10.13374/j.issn2095-9389.2018.s1.023

1. 北京科技大學冶金與生態工程學院;
2. 北京科技大學計算機與通信工程學院

基金項目:

國家重點研發計劃資助重點專項 (2016YFB0601301)

國家自然科學基金資助項目 (51574032)

中央高校基本科研業務費資助項目 (FRF-TP-16-081A1)

詳細信息

通訊作者:
馮凱, E-mail:fengkai-show@163.com

中圖分類號: TF769.4
計量
- 文章訪問數: 23
- HTML全文瀏覽量: 9
- PDF下載量: 0
- 被引次數: 0
出版歷程
- 收稿日期: 2018-01-20
- 網絡出版日期: 2023-07-18

End temperature prediction of molten steel in RH based on integrated case-based reasoning

1. School of Metallurgy and Ecology Engineering,University of Science and Technology Beijing;
2. School of Computer and Communication Engineering,University of Science and Technology Beijing

摘要

摘要: 針對RH工序終點鋼水溫度預測問題, 提出一種基于多元線性回歸和遺傳算法改進的集成案例推理方法.首先, 針對一般案例推理方法中缺少影響因素精選方法的問題, 利用多元線性回歸進行屬性約簡;然后, 針對案例檢索中相似度計算缺少權重計算方法的問題, 利用遺傳算法進行權重優化;最后, 基于精簡的影響因素和優化的權重, 利用改進灰色關聯相似度進行案例檢索, 實現RH終點鋼水溫度預測.利用某鋼鐵企業RH工序實際生產數據分別對多元線性回歸、BP神經網絡、一般案例推理方法和集成案例推理方法進行測試, 結果表明, 集成案例推理方法在多個溫度區間比多元線性回歸、BP神經網絡和一般案例推理方法都有更高的預測精度.
- 真空精煉 /
- 鋼水溫度 /
- 案例推理 /
- 預測 /
- 遺傳算法
Abstract: In regards to the end temperature prediction of molten steel in RH refining, an integrated case-based reasoning (CBR) method based on multiple linear regression (MLR) and genetic algorithm (GA) was proposed.Firstly, MLR was used to intelligently simplify the number of attributes to modify the lack of methods in the accurate selection of influencing factors in general CBR method.Secondly, GA was used to optimize the attribute weights in order to resolve the lack of attribute weights calculation method for similarity computation in case retrieval.Lastly, the end temperature prediction of molten steel in RH refining was realized based on the simplified influencing factors and optimized weights, and using grey relational degree (GRD) in case retrieval.Testing was performed based on the actual production data in RH refining in steelmaking plant, and comparison between MLR method, BP neural network, general CBR method and integrated CBR method was carried out.The results show that integrated CBR method has better prediction accuracy than MLR method, BP neural network and general CBR method in multiple temperature ranges.
- Ruhrstahl-Heraeus /
- molten steel temperature /
- case-based reasoning /
- prediction /
- genetic algorithm

HTML全文

參考文獻(16)

[1]	Wang H B, Cai J, Feng K. Predicting the endpoint phosphorus content of molten steel in BOF by two-stage hybrid method. J Iron Steel Res Int, 2014, 21 (S1) :65
[2]	Liang Y R, Wang H B, Xu A J, et al. A two-step case-based reasoning method based on attributes reduction for predicting the endpoint phosphorus content. ISIJ Int, 2015, 55 (5) :1035
[3]	Wang X, Yuan P, Mao Z, et al. Molten steel temperature prediction model based on bootstrap feature subsets ensemble regression trees. Knowledge Based Syst, 2016, 101 (6) :48
[4]	Wang X, You M, Mao Z, et al. Tree-structure ensemble general regression neural networks applied to predict the molten steel temperature in ladle furnace. Adv Eng Inf, 2016, 30 (3) :368
[5]	Yue Y J, Yao Y D, Zhao H, et al. BOF endpoint prediction based on multi-neural network model. Appl Mech Mater, 2014, 441:666
[6]	Wang Z, Xie F, Wang B, et al. The control and prediction of end-point phosphorus content during BOF steelmaking process.Steel Res Int, 2014, 85 (4) :599
[7]	Wang H B, Xu A J, Ai L X, et al. Prediction of endpoint phosphorus content of molten steel in BOF using weighted K-means and GMDH neural network. J Iron Steel Res Int, 2012, 19 (1) :11
[8]	Feng K, He D, Xu A, et al. End temperature prediction of molten steel in LF based on CBR-BBN. Steel Res Int, 2016, 87 (1) :79
[9]	LüW, Mao Z Z, Yuan P. Ladle furnace steel temperature prediction model based on partial linear regularization networks with sparse representation. Steel Res Int, 2012, 83 (3) :288
[10]	Ahmad I, Kano M, Hasebe S, et al. Gray-box modeling for prediction and control of molten steel temperature in tundish. J Process Control, 2014, 24 (4) :375
[11]	Okura T, Ahmad I, Kano M, et al. High-performance prediction of molten steel temperature in tundish through gray-box model.ISIJ Int, 2013, 53 (1) :76
[12]	Wang H B, Xu A J, Ai L X, et al. An integrated CBR model for predicting endpoint temperature of molten steel in AOD. ISIJ Int, 2012, 52 (1) :80
[13]	He F, Xu A J, Wang H B, et al. End temperature prediction of molten steel in LF based on CBR. Steel Res Int, 2012, 83 (11) :1079
[14]	Feng K, Wang H B, Xu A J, et al. Endpoint temperature prediction of molten steel in RH using improved case-based reasoning. Int J Miner Metall Mater, 2013, 20 (12) :1148
[15]	Pal S, Halder C. Optimization of phosphorous in steel produced by basic oxygen steel making process using multi-objective evolutionary and genetic algorithms. Steel Res Int, 2017, 88 (3) :art.No. 1600193
[16]	Worapradya K, Thanakijkasem P. Optimising steel production shedules via a hierarchical genetic algorithm. S Afr J Ind Eng, 2014, 25 (2) :209