A dynamic-inner LSTM prediction method for key alarm variables forecasting in chemical process

doi:10.1016/j.cjche.2022.08.024

Abstract

Abstract: With the increase in the complexity of industrial system, simply detecting and diagnosing a fault may be insufficient in some cases, and prognosing the fault ahead of time could have a certain necessity. Accurate prediction of key alarm variables in chemical process can indicate the possible change to reduce the probability of abnormal conditions. According to the characteristics of chemical process data, this work proposed a key alarm variables prediction model in chemical process based on dynamic-inner principal component analysis (DiPCA) and long short-term memory (LSTM). DiPCA is used to extract the most dynamic components for prediction. While LSTM is used to learn the relationship and predict the key alarm variables. This work used a simulation data set and a real hydrogenation process data set for applications and explained the model validity from the essential characteristics. Comparison of results with different models shows that our model has better prediction accuracy and performance, which can provide the basis for fault prognosis and health management.

Key words: Fault prognosis, Process systems, Safety, Prediction, Principal component analysis, Long short term memory

摘要： With the increase in the complexity of industrial system, simply detecting and diagnosing a fault may be insufficient in some cases, and prognosing the fault ahead of time could have a certain necessity. Accurate prediction of key alarm variables in chemical process can indicate the possible change to reduce the probability of abnormal conditions. According to the characteristics of chemical process data, this work proposed a key alarm variables prediction model in chemical process based on dynamic-inner principal component analysis (DiPCA) and long short-term memory (LSTM). DiPCA is used to extract the most dynamic components for prediction. While LSTM is used to learn the relationship and predict the key alarm variables. This work used a simulation data set and a real hydrogenation process data set for applications and explained the model validity from the essential characteristics. Comparison of results with different models shows that our model has better prediction accuracy and performance, which can provide the basis for fault prognosis and health management.

关键词: Fault prognosis, Process systems, Safety, Prediction, Principal component analysis, Long short term memory

Yiming Bai, Shuaiyu Xiang, Feifan Cheng, Jinsong Zhao. A dynamic-inner LSTM prediction method for key alarm variables forecasting in chemical process[J]. Chinese Journal of Chemical Engineering, 2023, 55(3): 266-276.

Yiming Bai, Shuaiyu Xiang, Feifan Cheng, Jinsong Zhao. A dynamic-inner LSTM prediction method for key alarm variables forecasting in chemical process[J]. 中国化学工程学报, 2023, 55(3): 266-276.

References

[1] P.A. Carson, C.J. Mumford, An analysis of incidents involving major hazards in the chemical industry, J. Hazard. Mater. 3 (2) (1979) 149–165.
[2] D. Crowl, J. Louvar, Chemical Process safety: Fundamentals with Applications, Pearson Education, 2001
[3] M. Madakyaru, F. Harrou, Y. Sun, Improved data-based fault detection strategy and application to distillation columns, Process. Saf. Environ. Prot. 107 (2017) 22–34.
[4] X.T. Bi, R.S. Qin, D.Y. Wu, S.D. Zheng, J.S. Zhao, One step forward for smart chemical process fault detection and diagnosis, Comput. Chem. Eng. 164 (2022) 107884.
[5] H. Wu, J.S. Zhao, Deep convolutional neural network model based chemical process fault diagnosis, Comput. Chem. Eng. 115 (2018) 185–197.
[6] S.D. Zheng, J.S. Zhao, A new unsupervised data mining method based on the stacked autoencoder for chemical process fault diagnosis, Comput. Chem. Eng. 135 (2020) 106755.
[7] X.T. Bi, J.S. Zhao, A novel orthogonal self-attentive variational autoencoder method for interpretable chemical process fault detection and identification, Process. Saf. Environ. Prot. 156 (2021) 581–597.
[8] H. Wu, J.S. Zhao, Fault detection and diagnosis based on transfer learning for multimode chemical processes, Comput. Chem. Eng. 135 (2020) 106731.
[9] D.Y. Wu, J.S. Zhao, Process topology convolutional network model for chemical process fault diagnosis, Process. Saf. Environ. Prot. 150 (2021) 93–109.
[10] B. Bhadriraju, J.S.I. Kwon, F. Khan, Risk-based fault prediction of chemical processes using operable adaptive sparse identification of systems (OASIS), Comput. Chem. Eng. 152 (2021) 107378.
[11] S.Y. Xiang, Y.M. Bai, J.S. Zhao, Medium-term prediction of key chemical process parameter trend with small data, Chem. Eng. Sci. 249 (2022) 117361.
[12] M.A.I. Khan, S. Imtiaz, F. Khan, Predictive warning system for nonlinear process plants, J. Process. Control 100 (2021) 1–10.
[13] W.D. Tian, M.G. Hu, C.K. Li, Fault prediction based on dynamic model and grey time series model in chemical processes, Chin. J. Chem. Eng. 22 (6) (2014) 643–650.
[14] K. Zhong, M. Han, B. Han, Data-driven based fault prognosis for industrial systems: a concise overview, IEEE/CAA J. Autom. Sin. 7 (2) (2020) 330–345.
[15] J.V. Kresta, J.F. Macgregor, T.E. Marlin, Multivariate statistical monitoring of process operating performance, Can. J. Chem. Eng. 69 (1) (1991) 35–47.
[16] B.C. Juricek, D.E. Seborg, W.E. Larimore, Predictive monitoring for abnormal situation management, J. Process. Control 11 (2) (2001) 111–128.
[17] C. Hamzaçebi, D. Akay, F. Kutay, Comparison of direct and iterative artificial neural network forecast approaches in multi-periodic time series forecasting, Expert Syst. Appl. 36 (2) (2009) 3839–3844.
[18] A.E. Pankratz, Forecasting with univariate Box-Jenkins models: concepts and cases, John Wiley & Sons. (1983)
[19] Ü.Ç. Büyükşahin, Ş. Ertekin, Improving forecasting accuracy of time series data using a new ARIMA-ANN hybrid method and empirical mode decomposition, Neurocomputing. 361 (2019) 151–163.
[20] X.F. Yuan, C. Ou, Y.L. Wang, C.H. Yang, W.H. Gui, A novel semi-supervised pre-training strategy for deep networks and its application for quality variable prediction in industrial processes, Chem. Eng. Sci. 217 (2020) 115509.
[21] J.T. Connor, R.D. Martin, L.E. Atlas, Recurrent neural networks and robust time series prediction, IEEE Trans. Neural Netw. 5 (2) (1994) 240–254.
[22] S. Hochreiter, J. Schmidhuber, Long short-term memory, Neural Comput 9 (8) (1997) 1735–1780.
[23] J.C. Song, L.Y. Zhang, G.X. Xue, Y.P. Ma, S. Gao, Q.L. Jiang, Predicting hourly heating load in a district heating system based on a hybrid CNN-LSTM model, Energy Build. 243 (2021) 110998.
[24] X. Zhang, Y.Y. Zou, S.Y. Li, S.H. Xu, A weighted auto regressive LSTM based approach for chemical processes modeling, Neurocomputing 367 (2019) 64–74.
[25] Z.X. Guo, J.Z. Zhao, Z.J. You, Y.M. Li, S. Zhang, Y.Y. Chen, Prediction of coalbed methane production based on deep learning, Energy 230 (2021) 120847.
[26] F.F. Cheng, Q.P. He, J.S. Zhao, A novel process monitoring approach based on variational recurrent autoencoder, Comput. Chem. Eng. 129 (2019) 106515.
[27] S.X. Ji, X.H. Han, Y.C. Hou, Y. Song, Q.F. Du, Remaining useful life prediction of airplane engine based on PCA–BLSTM, Sensors 20 (16) (2020) 4537.
[28] Huang, X. Yan, Dynamic process fault detection and diagnosis based on dynamic principal component analysis, dynamic independent component analysis and Bayesian inference, Chemometrics and Intelligent Laboratory Systems. 148 (2015) 115–127.
[29] W.F. Ku, R.H. Storer, C. Georgakis, Disturbance detection and isolation by dynamic principal component analysis, Chemom. Intell. Lab. Syst. 30 (1) (1995) 179–196.
[30] G. Li, B.S. Liu, S.J. Qin, D.H. Zhou, Dynamic latent variable modeling for statistical process monitoring, IFAC Proc. Vol. 44 (1) (2011) 12886–12891.
[31] Y.N. Dong, S.J. Qin, A novel dynamic PCA algorithm for dynamic data modeling and process monitoring, J. Process. Control 67 (2018) 1–11.
[32] A.V. Ooyen, B. Nienhuis, Improving the convergence of the back-propagation algorithm, Neural Netw. 5 (3) (1992) 465–471.
[33] Application of recurrent neural networks for drought projections in California, Atmospheric Research. 188 (2017) 100–106.
[34] Y. Bengio, P. Simard, P. Frasconi, Learning long-term dependencies with gradient descent is difficult, IEEE Trans Neural Netw 5 (2) (1994) 157–166.
[35] J. Forkman, J. Josse, H.P. Piepho, Hypothesis tests for principal component analysis when variables are standardized, J. Agric. Biol. Environ. Stat. 24 (2) (2019) 289–308.
[36] N.F. Thornhill, S.C. Patwardhan, S.L. Shah, A continuous stirred tank heater simulation model with applications, J. Process. Control 18 (3–4) (2008) 347–360.
[37] M.T. Amin, F. Khan, S. Ahmed, S. Imtiaz, A data-driven Bayesian network learning method for process fault diagnosis, Process. Saf. Environ. Prot. 150 (2021) 110–122.
[38] J. Yu, S.J. Qin, Multimode process monitoring with Bayesian inference-based finite Gaussian mixture models, Aiche J. 54 (7) (2008) 1811–1829.
[39] C.E. Shannon, A mathematical theory of communication, Bell Syst. Tech. J. 27 (4) (1948) 623–656.