A Data-Driven Online Multimodal Identification Method for Industrial Processes Based on Complex Network
Tóm tắt
With the increasing complexity of industrial processes, many practical industrial processes have multimodal characteristics to meet production requirements. In addition to stable modes with different operating points, transition modes are also generated when mode switching occurs. To monitor the running state of the process in real-time, it is necessary to carry out online multimodal identification. At present, how to accurately identify the transition modes and unknown modes is still an open problem. In this paper, an online multimodal identification method based on the complex network is proposed, which identifies different modes by the difference of data distribution between samples. To identify transition modes, moving window technology is introduced to extract dynamic characteristics of the historical data. The multimodal data is mapped in a complex network with the mean vectors of windows as nodes and the Jensen–Shannon (JS) divergence between the mean vectors as edges. Then, the community clustering algorithm is applied to cluster the nodes, which can automatically determine the number of clustering and solve the problem of unknown modal numbers in historical data. On this basis, the Kernel Density Estimation method is used to introduce the JS divergence thresholds to realize a new online multimodal identification, including transition modes and unknown modes. The model can be updated by collecting samples of unknown modes from the online multimodal identification results. Finally, the effectiveness of the proposed method is verified by a numerical example and the Tennessee-Eastman process.
Tài liệu tham khảo
Bathelt, A., Ricker, N. L., & Jelali, M. (2015). Revision of the tennessee eastman process model. IFAC-PapersOnLine, 48(8), 309–314.
Blondel, V. D., Guillaume, J.-L., Lambiotte, R., & Lefebvre, E. (2008). Fast unfolding of communities in large networks. Journal of statistical mechanics: Theory and experiment, 2008(10), P10008.
Chen, S., Jiang, Q., & Yan, X. (2020). Multimodal process monitoring based on transition-constrained gaussian mixture model. Chinese Journal of Chemical Engineering, 28(12), 3070–3078.
Downs, J. J., & Vogel, E. F. (1993). A plant-wide industrial process control problem. Computers & Chemical Engineering, 17(3), 245–255.
Gao, H., Wei, C., Huang, W., & Gao, X. (2022). Multimode process monitoring based on hierarchical mode identification and stacked denoising autoencoder. Chemical Engineering Science, 253, 117556.
Lin, J. (1991). Divergence measures based on the shannon entropy. IEEE Transactions on Information theory, 37(1), 145–151.
Newman, M. E. (2006). Modularity and community structure in networks. Proceedings of the National Academy of Sciences, 103(23), 8577–8582.
Quiñones-Grueiro, M., Prieto-Moreno, A., Verde, C., & Llanes-Santiago, O. (2019). Data-driven monitoring of multimode continuous processes: A review. Chemometrics and Intelligent Laboratory Systems, 189, 56–71.
Rashid, M. M., & Yu, J. (2012). Hidden markov model based adaptive independent component analysis approach for complex chemical process monitoring and fault detection. Industrial & Engineering Chemistry Research, 51(15), 5506–5514.
Ricker, N. (1995). Optimal steady-state operation of the tennessee eastman challenge process. Computers & Chemical Engineering, 19(9), 949–959.
Sales-Pardo, M., Guimera, R., Moreira, A. A., & Amaral, L. A. N. (2007). Extracting the hierarchical organization of complex systems. Proceedings of the National Academy of Sciences, 104(39), 15224–15229.
Scott, D. W. (2015). Multivariate density estimation: Theory, practice, and visualization. John Wiley & Sons.
Sun, Y.-N., Zhuang, Z.-L., Xu, H.-W., Qin, W., & Feng, M.-J. (2021). Data-driven modeling and analysis based on complex network for multimode recognition of industrial processes, Journal of Manufacturing Systems.
Sun, B., Yang, C., Wang, Y., Gui, W., Craig, I., & Olivier, L. (2020). A comprehensive hybrid first principles/machine learning modeling framework for complex industrial processes. Journal of Process Control, 86, 30–43.
Wang, L., Yang, C., & Sun, Y. (2018). Multimode process monitoring approach based on moving window hidden markov model. Industrial & Engineering Chemistry Research, 57(1), 292–301.
Wu, D., Zhou, D., Zhang, J., & Chen, M. (2020). Multimode process monitoring based on fault dependent variable selection and moving window-negative log likelihood probability. Computers & Chemical Engineering, 136, 106787.