A journal of IEEE and CAA , publishes high-quality papers in English on original theoretical/experimental research and development in all areas of automation
Volume 9 Issue 4
Apr.  2022

IEEE/CAA Journal of Automatica Sinica

  • JCR Impact Factor: 11.8, Top 4% (SCI Q1)
    CiteScore: 23.5, Top 2% (Q1)
    Google Scholar h5-index: 77, TOP 5
Turn off MathJax
Article Contents
Z. L. Yuan, X. R. Li, D. Wu, X. J. Ban, N.-Q. Wu, H.-N. Dai, and H. Wang, “Continuous-time prediction of industrial paste thickener system with differential ODE-net,” IEEE/CAA J. Autom. Sinica, vol. 9, no. 4, pp. 686–698, Apr. 2022. doi: 10.1109/JAS.2022.105464
Citation: Z. L. Yuan, X. R. Li, D. Wu, X. J. Ban, N.-Q. Wu, H.-N. Dai, and H. Wang, “Continuous-time prediction of industrial paste thickener system with differential ODE-net,” IEEE/CAA J. Autom. Sinica, vol. 9, no. 4, pp. 686–698, Apr. 2022. doi: 10.1109/JAS.2022.105464

Continuous-Time Prediction of Industrial Paste Thickener System With Differential ODE-Net

doi: 10.1109/JAS.2022.105464
Funds:  This work was supported by National Key Research and Development Program of China (2019YFC0605300), the National Natural Science Foundation of China (61873299, 61902022, 61972028), Scientific and Technological Innovation Foundation of Shunde Graduate School, University of Science and Technology Beijing (BK21BF002), Macao Science and Technology Development Fund under Macao Funding Scheme for Key R&D Projects (0025/2019/AKP), and Macao Science and Technology Development Fund (0015/2020/AMJ)
More Information
  • It is crucial to predict the outputs of a thickening system, including the underflow concentration (UC) and mud pressure, for optimal control of the process. The proliferation of industrial sensors and the availability of thickening-system data make this possible. However, the unique properties of thickening systems, such as the non-linearities, long-time delays, partially observed data, and continuous time evolution pose challenges on building data-driven predictive models. To address the above challenges, we establish an integrated, deep-learning, continuous time network structure that consists of a sequential encoder, a state decoder, and a derivative module to learn the deterministic state space model from thickening systems. Using a case study, we examine our methods with a tailing thickener manufactured by the FLSmidth installed with massive sensors and obtain extensive experimental results. The results demonstrate that the proposed continuous-time model with the sequential encoder achieves better prediction performances than the existing discrete-time models and reduces the negative effects from long time delays by extracting features from historical system trajectories. The proposed method also demonstrates outstanding performances for both short and long term prediction tasks with the two proposed derivative types.

     

  • loading
  • [1]
    S. Yin, Y. Shao, A. Wu, H. Wang, X. Liu, and Y. Wang, “A systematic review of paste technology in metal mines for cleaner production in china,” Journal of Cleaner Production, vol. 247, Article No. 119590, 2020. doi: 10.1016/j.jclepro.2019.119590
    [2]
    Z.-L. Yuan, R.-Z. He, C. Yao, J. Li, X.-J. Ban, and X.-R. Li, “Online reinforcement learning control algorithm for concentration of thickener underflow,” Acta Automatica Sinica, vol. 45, pp. 1–15, 2019.
    [3]
    H. Li, A. Wu, and H. Wang, “Evaluation of short-term strength development of cemented backfill with varying sulphide contents and the use of additives,” Journal of Environmental Management, vol. 239, pp. 279–286, 2019.
    [4]
    C. K. Tan, J. Bao, and G. Bickert, “A study on model predictive control in paste thickeners with rake torque constraint,” Minerals Engineering, vol. 105, pp. 52–62, 2017. doi: 10.1016/j.mineng.2017.01.011
    [5]
    F. Núnez, S. Langarica, Díaz, M. Torres, and J. C. Salas, “Neural network-based model predictive control of a paste thickener over an industrial internet platform,” IEEE Trans. Industrial Informatics, vol. 16, no. 4, pp. 2859–2867, 2019.
    [6]
    Z. Yuan, J. Hu, D. Wu, and X. Ban, “A dual-attention recurrent neural network method for deep cone thickener underflow concentration prediction,” Sensors (Switzerland), vol. 20, no. 5, pp. 1–18, 2020.
    [7]
    A. Wu, Z. Ruan, R. Bürger, S. Yin, J. Wang, and Y. Wang, “Optimization of flocculation and settling parameters of tailings slurry by response surface methodology,” Minerals Engineering, vol. 156, Article No. 106488, 2020. doi: 10.1016/j.mineng.2020.106488
    [8]
    E. K. Larsson and T. Söderström, “Identification of continuous-time AR processes from unevenly sampled data,” Automatica, vol. 38, no. 4, pp. 709–718, 2002. doi: 10.1016/S0005-1098(01)00244-8
    [9]
    A. Voulodimos, N. Doulamis, A. Doulamis, and E. Protopapadakis, “Deep learning for computer vision: A brief review,” Computational Intelligence and Neuroscience, vol. 2018, PP. 1–13, 2018.
    [10]
    B. Ma, X. Wei, C. Liu, X. Ban, H. Huang, H. Wang, W. Xue, S. Wu, M. Gao, Q. Shen, “Data augmentation in microscopic images for material data mining,” NPJ Computational Materials, vol. 6, no. 1, pp. 1–9, 2020. doi: 10.1038/s41524-019-0267-z
    [11]
    B. Ma, Y. Zhu, X. Yin, X. Ban, H. Huang, and M. Mukeshimana, “Sesffuse: An unsupervised deep model for multi-focus image fusion,” Neural Computing and Applications, vol. 33, no. 11, pp. 5793–5804, 2021. doi: 10.1007/s00521-020-05358-9
    [12]
    T. Young, D. Hazarika, S. Poria, and E. Cambria, “Recent trends in deep learning based natural language processing,” IEEE Computational Intelligence Magazine, vol. 13, no. 3, pp. 55–75, 2018. doi: 10.1109/MCI.2018.2840738
    [13]
    H. Liu, I. Chatterjee, M. Zhou, X. S. Lu, and A. Abusorrah, “Aspectbased sentiment analysis: A survey of deep learning methods,” IEEE Trans. Computational Social Systems, vol. 7, no. 6, pp. 1358–1375, 2020. doi: 10.1109/TCSS.2020.3033302
    [14]
    A. E. Essien and C. Giannetti, “A deep learning model for smart manufacturing using convolutional LSTM neural network autoencoders,” IEEE Trans. Industrial Informatics, vol. 16, no. 9, pp. 6069–6078, 2020. doi: 10.1109/TII.2020.2967556
    [15]
    J. Liu, N. Wu, Y. Qiao, and Z. Li, “Short-term traffic flow forecasting using ensemble approach based on deep belief networks,” IEEE Trans. Intelligent Transportation Systems, pp. 1–14, 2020. DOI: 10.1109/TITS.2020.3011700
    [16]
    J. Fei and L. Liu, “Real-time nonlinear model predictive control of active power filter using self-feedback recurrent fuzzy neural network estimator,” IEEE Trans. Industrial Electronics, pp. 1–1, 2021. DOI: 10.1109/TIE.2021.3106007
    [17]
    D. A. Neu, J. Lahann, and Fettke, “A systematic literature review on state-of-the-art deep learning methods for process prediction,” Artificial Intelligence Review, pp. 1–27, 2021. DOI: 10.1007/s10462-021-09960-8
    [18]
    H. Li, G. Hu, J. Li, and M. Zhou, “Intelligent fault diagnosis for large-scale rotating machines using binarized deep neural networks and random forests,” IEEE Trans. Automation Science and Engineering, pp. 1–11, 2021. DOI: 10.1109/TASE.2020.3048056
    [19]
    T. Demeester, “System identification with time-aware neural sequence models,” in Proc. AAAI Conf. Artificial Intelligence, vol. 34, no. 4, 2020, pp. 3757–3764.
    [20]
    C. K. Tan, R. Setiawan, J. Bao, and G. Bickert, “Studies on parameter estimation and model predictive control of paste thickeners,” Journal of Process Control, vol. 28, pp. 1–8, 2015. doi: 10.1016/j.jprocont.2015.02.002
    [21]
    J. I. Langlois and A. Cipriano, “Dynamic modeling and simulation of tailing thickener units for the development of control strategies,” Minerals Engineering, vol. 131, pp. 131–139, 2019. doi: 10.1016/j.mineng.2018.11.006
    [22]
    X. Luo, M. Zhou, S. Li, and M. Shang, “An inherently nonnegative latent factor model for high-dimensional and sparse matrices from industrial applications,” IEEE Trans. Industrial Informatics, vol. 14, no. 5, pp. 2011–2022, 2017.
    [23]
    X. Luo, M. Zhou, S. Li, Y. Xia, Z.-H. You, Q. Zhu, and H. Leung, “Incorporation of efficient second-order solvers into latent factor models for accurate prediction of missing qos data,” IEEE Transactions on Cybernetics, vol. 48, no. 4, pp. 1216–1228, 2017.
    [24]
    D. Wu, X. Luo, M. Shang, Y. He, G. Wang, and X. Wu, “A datacharacteristic-aware latent factor model for web services QoS prediction,” IEEE Trans. Knowledge and Data Engineering, 2020.
    [25]
    M. Kebria, A. Khosravi, S. M. Salaken, and S. Nahavandi, “Deep imitation learning for autonomous vehicles based on convolutional neural networks,” IEEE/CAA Journal of Automatica Sinica, vol. 7, no. 1, pp. 82–95, 2020. doi: 10.1109/JAS.2019.1911825
    [26]
    W. Liu, Z. Wang, X. Liu, N. Zeng, Y. Liu, and F. E. Alsaadi, “A survey of deep neural network architectures and their applications,” Neurocomputing, vol. 234, pp. 11–26, 2017. doi: 10.1016/j.neucom.2016.12.038
    [27]
    T.-Y. Chai, “Development directions of industrial artificial intelligence,” Acta Automatica Sinica, vol. 46, no. 10, pp. 2005–2012, 2020.
    [28]
    L. Jay, L. Xiang, X. Yuan-Ming, Y. Shaojie, and S. Ke-Yi, “Recent advances and prospects in industrial AI and ap-plications,” Acta Automatica Sinica, vol. 46, no. 10, pp. 2031–2044, 2020.
    [29]
    D. Wu, H. Wang, and R. Seidu, “Collaborative analysis for computational risk in urban water supply systems,” in Proc. 28th ACM Int. Conf. Information and Knowledge Management, 2019, pp. 2297–2300.
    [30]
    D. Wu, H. Wang, H. Mohammed, and R. Seidu, “Quality risk analysis for sustainable smart water supply using data perception,” IEEE Trans. Sustainable Computing, vol. 5, no. 3, pp. 377–388, 2020. doi: 10.1109/TSUSC.2019.2929953
    [31]
    J. Zhou, H.-N. Dai, and H. Wang, “Lightweight convolution neural networks for mobile edge computing in transportation cyber physical systems,” ACM Trans. Intelligent Systems and Technology (TIST), vol. 10, no. 6, pp. 1–20, 2019.
    [32]
    H.-N. Dai, R. C.-W. Wong, H. Wang, Z. Zheng, and A. V. Vasilakos, “Big data analytics for large-scale wireless networks: Challenges and opportunities,” ACM Comput. Surv., vol. 52, no. 5, pp. 99:1–99:36, 2019.
    [33]
    R. Oulhiq, K. Benjelloun, Y. Kali, and M. Saad, “Identification and control of an industrial thickener using historical data,” in Proc. 18th Int. Multi-Conf. on Systems, Signals Devices (SSD), 2021, pp. 915–920.
    [34]
    Di az, J. C. Salas, A. Cipriano, and F. Núnez, “Random forest model? Predictive control for paste thickening” Minerals Engineering, vol. 163, Article No. 106760, 2021. doi: 10.1016/j.mineng.2020.106760
    [35]
    R. T. Q. Chen, Y. Rubanova, J. Bettencourt, and D. K. Duvenaud, “Neural ordinary differential equations,” in Proc. Advances in Neural Information Processing Systems 31, Curran Associates, Inc., 2018, pp. 6571–6583.
    [36]
    R. J. Weiss, J. Chorowski, N. Jaitly, Y. Wu, and Z. Chen, “Sequenceto-sequence models can directly translate foreign speech,” in Proc. Interspeech 2017, pp. 2625–2629, 2017.
    [37]
    M. GEVERS, “A personal view of the development of system identification: A 30-year journey through an exciting field,” IEEE Control Systems Magazine, vol. 26, no. 6, pp. 93–105, 2006. doi: 10.1109/MCS.2006.252834
    [38]
    T. Chai, Y. Jia, H. Li, and H. Wang, “An intelligent switching control for a mixed separation thickener process,” Control Engineering Practice, vol. 57, pp. 61–71, 2016. doi: 10.1016/j.conengprac.2016.07.007
    [39]
    B. Kim and M. S. Klima, “Development and application of a dynamic model for hindered-settling column separations,” Minerals Engineering, vol. 17, no. 3, pp. 403–410, 2004. doi: 10.1016/j.mineng.2003.11.013
    [40]
    M. Guidolin and M. Pedio, “Chapter 4-unit roots and cointegration,” in Essentials of Time Series for Financial Applications, M. Guidolin and M. Pedio, Eds. USA: Academic Press, 2018, pp. 113–149.
    [41]
    F. Christoffersen, “Forecasting non-stationary economic time series,” Journal of the American Statistical Association, vol. 96, no. 453, pp. 347–347, 2001.
    [42]
    D. P. Kingma and J. Ba, “Adam: A method for stochastic optimization,” arXiv preprint arXiv: 1412.6980, 2014.
    [43]
    Bogacki and L. F. Shampine, “A 3(2) pair of Runge-Kutta formulas,” Applied Mathematics Letters, vol. 2, no. 4, pp. 321–325, 1989. doi: 10.1016/0893-9659(89)90079-7
    [44]
    S. Wu, X. Xiao, Q. Ding, P. Zhao, Y. Wei, and J. Huang, “Adversarial sparse transformer for time series forecasting,” Advances in Neural Information Processing Systems, vol. 33, no. NeurIPS, pp. 17105–17115, 2020.
    [45]
    S. S. Rangapuram, M. Seeger, J. Gasthaus, L. Stella, Y. Wang, and T. Januschowski, “Deep state space models for time series forecasting,” Advances in Neural Information Processing Systems, vol. 2018-Decem, no. NeurIPS, pp. 7785–7794, 2018.
    [46]
    H. Lu, L. Jin, X. Luo, B. Liao, D. Guo, and L. Xiao, “RNN for solving perturbed time-varying underdetermined linear system with double bound limits on residual errors and state variables,” IEEE Trans. Industrial Informatics, vol. 15, no. 11, pp. 5931–5942, 2019. doi: 10.1109/TII.2019.2909142

Catalog

    通讯作者: 陈斌, bchen63@163.com
    • 1. 

      沈阳化工大学材料科学与工程学院 沈阳 110142

    1. 本站搜索
    2. 百度学术搜索
    3. 万方数据库搜索
    4. CNKI搜索

    Figures(7)  / Tables(5)

    Article Metrics

    Article views (887) PDF downloads(112) Cited by()

    Highlights

    • Networks based on ODE-net perform better in continuous-time system prediction.
    • Stationary ODE-net outperforms the non-stationary ones in long-term prediction.
    • The ODE-nets with high-order solvers have better accuracy.
    • The time-delay in predicting thickening system is about 80min.

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return