0
Journal of Manufacturing Science and Engineering | Volume 138 | Issue 10 | research-article
Research Papers

A Data-Driven Model for Energy Consumption in the Sintering Process

[+] Author and Article Information
Junkai Wang

School of Electronics and
Information Engineering,
Tongji University,
Shanghai 201804, China

Fei Qiao

School of Electronics and
Information Engineering,
Tongji University,
4800 Cao'an Road,
Jiading District,
Shanghai 201804, China
e-mail: fqiao@tongji.edu.cn

Fu Zhao

School of Mechanical Engineering,
Purdue University,
West Lafayette, IN 47907;Environmental and Ecological Engineering,
Purdue University,
West Lafayette, IN 47907

John W. Sutherland

Environmental and Ecological Engineering,
Purdue University,
West Lafayette, IN 47907

1Corresponding author.

Manuscript received December 2, 2015; final manuscript received May 11, 2016; published online June 22, 2016. Editor: Y. Lawrence Yao.

J. Manuf. Sci. Eng 138(10), 101001 (Jun 22, 2016) (12 pages) Paper No: MANU-15-1632; doi: 10.1115/1.4033661 History: Received December 02, 2015; Revised May 11, 2016
Article
References
Figures
Tables

As environmental performance becomes increasingly important, the sintering process is receiving more attention since it consumes large amounts of energy. This paper proposes a data-driven model for sintering energy consumption, which considers both model accuracy and time efficiency. The proposed model begins with removing data anomalies using a local outlier factor (LOF) algorithm and an attribute selection module using the RReliefF method. Then, to accurately predict sintering energy consumption, an integrated predictive model is employed that uses bagging-enhanced extreme learning machine (ELM) and support vector regression (SVR) machine, combined with an entropy weight method. A case study is used to demonstrate the effectiveness of the proposed model using actual production data for a year. Results show that the proposed model outperforms other models and is computationally efficient. Optimal parameters of the LOF (1.3) and number of attributes (30) were identified. It was found that coke powder has the most significant impact on the solid energy consumption (SEC), while cooling water flow rate provides the most significant impact on the gas energy consumption (GEC) within each recorded attribute variation. Parametric analysis further revealed the relationships between energy consumption and the significant attributes mentioned above. It is suggested that the proposed model could effectively reduce the energy consumption by attaining more efficient attribute settings.
FIGURES IN THIS ARTICLE
<>
Copyright © 2016 by ASME
Your Session has timed out. Please sign back in to continue.

References

Figures

Grahic Jump Location
Fig. 1

Diagram of sintering process (adapted from Ref. [6])

+Diagram of sintering process (adapted from Ref. [6])
View Large  |  Save Figure  |  Download Slide (.ppt)
Diagram of sintering process (adapted from Ref. [6])
Grahic Jump Location
Fig. 2

Framework of the sintering energy consumption integrated predictive model

+Framework of the sintering energy consumption integrated predictive model
View Large  |  Save Figure  |  Download Slide (.ppt)
Framework of the sintering energy consumption integrated predictive model
Grahic Jump Location
Fig. 3

Predicted values and measured values of testing sintering sample. For SEC (in kg/t) and GEC (in m3/t), circles denote predicted values, and asterisks denote measured values, respectively.

+Predicted values and measured values of testing sintering sample. For SEC (in kg/t) and GEC (in m3/t), circles denote predicted values, and asterisks denote measured values, respectively.
View Large  |  Save Figure  |  Download Slide (.ppt)
Predicted values and measured values of testing sintering sample. For SEC (in kg/t) and GEC (in m3/t), circles denote predicted values, and asterisks denote measured values, respectively.
Grahic Jump Location
Fig. 4

Performance of SEC prediction using different models (θ=1.3 and η=30)

+Performance of SEC prediction using different models (θ=1.3 and η=30)
View Large  |  Save Figure  |  Download Slide (.ppt)
Performance of SEC prediction using different models (θ=1.3 and η=30)
Grahic Jump Location
Fig. 5

Performance of GEC prediction using different models (θ=1.3 and η=30)

+Performance of GEC prediction using different models (θ=1.3 and η=30)
View Large  |  Save Figure  |  Download Slide (.ppt)
Performance of GEC prediction using different models (θ=1.3 and η=30)
Grahic Jump Location
Fig. 6

Performance under different model schemas for SEC. For “I.M.LOF,” θ is set to 1.3; for “I.M.RReliefF,” η is set to 30; and for “I.M.L.R,”  θ=1.3  and  η=30. The results are derived from averaging 30 runs.

+Performance under different model schemas for SEC. For “I.M.LOF,” θ is set to 1.3; for “I.M.RReliefF,” η is set to 30; and for “I.M.L.R,”  θ=1.3  and  η=30. The results are derived from averaging 30 runs.
View Large  |  Save Figure  |  Download Slide (.ppt)
Performance under different model schemas for SEC. For “I.M.LOF,” θ is set to 1.3; for “I.M.RReliefF,” η is set to 30; and for “I.M.L.R,”  θ=1.3  and  η=30. The results are derived from averaging 30 runs.
Grahic Jump Location
Fig. 7

Performance of the proposed model with respect to η under different θ for SEC. All the results are calculated by averaging 30 runs.

+Performance of the proposed model with respect to η under different θ for SEC. All the results are calculated by averaging 30 runs.
View Large  |  Save Figure  |  Download Slide (.ppt)
Performance of the proposed model with respect to η under different θ for SEC. All the results are calculated by averaging 30 runs.
Grahic Jump Location
Fig. 8

Sensitivity analysis of the SEC and GEC over the top 30 attributes (Nos. 1–10 denote the attributes corresponding to the ranking in Tables 3 and 4, respectively)

+Sensitivity analysis of the SEC and GEC over the top 30 attributes (Nos. 1–10 denote the attributes corresponding to the ranking in Tables 3 and 4, respectively)
View Large  |  Save Figure  |  Download Slide (.ppt)
Sensitivity analysis of the SEC and GEC over the top 30 attributes (Nos. 1–10 denote the attributes corresponding to the ranking in Tables 3 and 4, respectively)
Grahic Jump Location
Fig. 9

Parametric analyses of the proposed model. SEC is in kg/t, GEC in m3/t, coke in t, and cooling water in t.

+Parametric analyses of the proposed model. SEC is in kg/t, GEC in m3/t, coke in t, and cooling water in t.
View Large  |  Save Figure  |  Download Slide (.ppt)
Parametric analyses of the proposed model. SEC is in kg/t, GEC in m3/t, coke in t, and cooling water in t.

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Design of Energy Efficient Smart Home Appliance to Run on Low DC Voltage for Reducing Power Consumption
International Conference on Software Technology and Engineering, 3rd (ICSTE 2011)
Overview
Energy, Technology, and the Environment> Chapter 1
Energy Sources and Supplies
Energy Supply and Pipeline Transportation: Challenges & Opportunities> Chapter 2
Dynamic Control Strategy of a Biped Inspired from Human Walking
Intelligent Engineering Systems through Artificial Neural Networks Volume 18
Analysis of Energy Consumption for Water Source Heat Pump System in Wuhan, China
Inaugural US-EU-China Thermophysics Conference-Renewable Energy 2009 (UECTC 2009 Proceedings)
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In