A Time-Series Motif-Based Rule Fusion Method for Interpretable State Prediction of Aluminum Electrolysis Cells

Danyang Cao; Shuobo Yu; Zifeng Lin; Gao Lei

doi:10.54097/r9rfjp62

Authors

Danyang Cao
Shuobo Yu
Zifeng Lin
Gao Lei

DOI:

https://doi.org/10.54097/r9rfjp62

Keywords:

Aluminum electrolysis, Process monitoring, State prediction, Rule mining, Motif discovery, Interpretable prediction

Abstract

Accurate state prediction is important for improving the stability and efficiency of aluminum electrolysis cells, but practical prediction remains difficult because the production process contains strongly coupled variables, delayed process responses, and many decisions based on operator experience. To address these issues, this paper presents an interpretable motif-based rule fusion method for cell state prediction. The method first separates industrial variables into decision variables and state variables so that manual control actions and automatically monitored responses can be modeled with clear semantic roles. Multivariate motif groups are then mined from each dataset, after which temporal rules and association rules are extracted. Association rules describe delayed operation-response relationships between decision motifs and subsequent state motifs, while temporal rules describe weekday and monthly occurrence regularities of state motifs. During prediction, the method prioritizes matched association rules and uses temporal rules as a fallback when no reliable association rule is available. Experiments on large-scale data from 3,391 pot controllers show that the combined rule-fusion strategy achieves lower pattern-level prediction error than single-rule strategies and representative forecasting baselines. The method also provides explicit evidence linking interventions to later state evolution, supporting interpretable process monitoring and operator-oriented decision support.

Downloads

Download data is not yet available.

References

[1] Sun, Y., Chen, X., Cen, L., et al. (2024). A dynamic graph structure identification method of spatio-temporal correlation in an aluminum electrolysis cell. Applied Soft Computing, 157, 111536. https://doi.org/10.1016/j.asoc.2024.111536

[2] Zhou, L., Yao, Z., Sun, K., et al. (2024). Methodological review of methods and technology for utilization of spent carbon cathode in aluminum electrolysis. Energies, 17(19), 4866. https://doi.org/10.3390/en17194866

[3] Wang, J., Xie, Y., Xie, S., et al. (2024). Development of data-knowledge-driven predictive model and multi-objective optimization for intelligent optimal control of aluminum electrolysis process. Engineering Applications of Artificial Intelligence, 134, 108664. https://doi.org/10.1016/j.engappai.2024.108664

[4] Cao, D., & Lin, Z. (2024). Multidimensional time series motif group discovery based on matrix profile. Knowledge-Based Systems, 304, 112509. https://doi.org/10.1016/j.knosys.2024.112509

[5] Lin, J., Keogh, E., Lonardi, S., & Chiu, B. (2002). Finding motifs in time series. In Proceedings of the 2nd Workshop on Temporal Data Mining (pp. 23–26).

[6] Pradhan, G. N., & Prabhakaran, B. (2017). Association rule mining in multiple, multidimensional time series medical data. Journal of Healthcare Informatics Research, 1(1), 92–118. https://doi.org/10.1007/s41666-017-0008-6

[7] Yi, J., Bai, J., Zhou, W., et al. (2018). Operating parameters optimization for the aluminum electrolysis process using an improved quantum-behaved particle swarm algorithm. IEEE Transactions on Industrial Informatics, 14(8), 3405–3415. https://doi.org/10.1109/TII.2018.2820099

[8] Wang, J., Xie, Y., Xie, S., et al. (2024). Operational decision-making optimization of aluminum electrolysis process based on health evaluation of aluminum electrolytic cell. In 2024 IEEE International Conference on Cybernetics and Intelligent Systems and IEEE International Conference on Robotics, Automation and Mechatronics (pp. 156–161). https://doi.org/10.1109/CIS-RAM61216.2024.10634521

[9] Wang, J., Xie, S., Xie, Y., et al. (2024). A general knowledge-guided framework based on deep probabilistic network for enhancing industrial process modeling. IEEE Transactions on Industrial Informatics, 20(3), 3050–3059. https://doi.org/10.1109/TII.2023.3338896

[10] Berndt, D. J., & Clifford, J. (1994). Using dynamic time warping to find patterns in time series. In Proceedings of the 3rd International Conference on Knowledge Discovery and Data Mining (pp. 359–370).

[11] Keogh, E., Chakrabarti, K., Pazzani, M., & Mehrotra, S. (2001). Locally adaptive dimensionality reduction for indexing large time series databases. In Proceedings of the 2001 ACM SIGMOD International Conference on Management of Data (pp. 151–162). https://doi.org/10.1145/375663.375679

[12] Agrawal, R., Imieliński, T., & Swami, A. (1993). Mining association rules between sets of items in large databases. In Proceedings of the 1993 ACM SIGMOD International Conference on Management of Data (pp. 207–216). https://doi.org/10.1145/170035.170072

[13] Zaki, M. J. (2001). SPADE: An efficient algorithm for mining frequent sequences. Machine Learning, 42(1), 31–60. https://doi.org/10.1023/A:1007638616156

[14] Pei, J., Han, J., Mortazavi-Asl, B., et al. (2001). PrefixSpan: Mining sequential patterns efficiently by prefix-projected pattern growth. In Proceedings of the 17th International Conference on Data Engineering (pp. 215–224). https://doi.org/10.1109/ICDE.2001.914830

[15] Srikant, R., & Agrawal, R. (1996). Mining sequential patterns: Generalizations and performance improvements. In International Conference on Extending Database Technology (pp. 1–17). https://doi.org/10.1007/BFb0054687

[16] Das, G., Lin, K. I., Mannila, H., et al. (1998). Rule discovery from time series. In Proceedings of the Fourth International Conference on Knowledge Discovery and Data Mining (Vol. 98, pp. 16–22).

[17] Harms, S. K., Deogun, J., & Tadesse, T. (2002). Discovering sequential association rules with constraints and time lags in multiple sequences. In International Symposium on Methodologies for Intelligent Systems (pp. 432–441). https://doi.org/10.1007/3-540-45813-1_41

[18] Hoeppner, F. (2002). Learning dependencies in multivariate time series. In Proceedings of the ECAI (Vol. 2, pp. 25–31).

[19] Sacchi, L., Larizza, C., Combi, C., & Pozzi, G. (2007). Data mining with temporal abstractions: Learning rules from time series. Data Mining and Knowledge Discovery, 15(2), 217–247. https://doi.org/10.1007/s10618-007-0066-7

[20] Wu, H., Hu, T., Liu, Y., et al. (2023). TimesNet: Temporal 2D-variation modeling for general time series analysis. In International Conference on Learning Representations.

[21] Wu, H., Xu, J., Wang, J., et al. (2021). Autoformer: Decomposition transformers with auto-correlation for long-term series forecasting. Advances in Neural Information Processing Systems, 34, 22419–22430.

[22] Zeng, A., Chen, M., Zhang, L., & Xu, Q. (2023). Are transformers effective for time series forecasting? In Proceedings of the AAAI Conference on Artificial Intelligence (Vol. 37, No. 9, pp. 11121–11128). https://doi.org/10.1609/aaai.v37i9.26333

[23] Zhou, T., Ma, Z., Wen, Q., et al. (2022). FEDformer: Frequency enhanced decomposed transformer for long-term series forecasting. In International Conference on Machine Learning (pp. 27268–27286).

[24] Zhou, H., Zhang, S., Peng, J., et al. (2021). Informer: Beyond efficient transformer for long sequence time-series forecasting. In Proceedings of the AAAI Conference on Artificial Intelligence (Vol. 35, No. 12, pp. 11106–11115). https://doi.org/10.1609/aaai.v35i12.17325

[25] Campos, D., Zhang, M., Yang, B., et al. (2023). LightTS: Lightweight time series classification with adaptive ensemble distillation. Proceedings of the ACM on Management of Data, 1(2), 1–27. https://doi.org/10.1145/3588772.3594269

[26] Van Wesenbeeck, D., Yurtman, A., Meert, W., & Van den Berghe, S. (2024). LoCoMotif: Discovering time-warped motifs in time series. Data Mining and Knowledge Discovery, 38(4), 2276–2305. https://doi.org/10.1007/s10618-024-01030-9

[27] Tang, C., Xu, L., Yang, B., et al. (2023). GRU-based interpretable multivariate time series anomaly detection in industrial control system. Computers & Security, 127, 103094. https://doi.org/10.1016/j.cose.2023.103094

[28] Rizzi, W., Comuzzi, M., Di Francescomarino, C., & Maggi, F. M. (2024). Explainable predictive process monitoring: A user evaluation. Process Science, 1(1), 3. https://doi.org/10.1007/s44388-024-00003-8