Industrial process time-series modeling based on adapted receptive field temporal convolution networks concerning multi-region operations

Abstract Traditionally, recurrent neural networks (RNNs) are employed to deal with industrial process data with time-series characteristics. However, the time-consuming iterative feature of RNN may result in slow computation speeds and limited modeling accuracies when dealing with high dimensional and long horizon data from industrial operation. In response to this problem, an adapted receptive field temporal convolution network concerning multi-region operations (MRO-ARFTCN) is introduced in this paper. The MRO-ARFTCN algorithm is a variant of convolutional neural networks (CNNs), which can both extract causal characteristics among data and inherit forwarding propagations of CNNs. Specifically, multi-region operations characteristics are extracted from industrial process historical data using clustering approaches before delivering to networks. An adapted receptive field is responsible for storing local data and multi-region operations information, in which, local data can be adjusted to an appropriate dimension according to the dilated factor of the network, which greatly improves the time-series modeling performance. To verify the effectiveness of the proposed method, an industrial methanol production process is considered, which demonstrates a satisfying fitting result. Additionally, compared with conventional RNNs and long-short-term-memory (LSTM) methods, the MRO-ARFTCN has been proved more tractable and useful for industrial process time-series modeling.

[1]  Kwok-Wing Chau,et al.  Data-driven models for monthly streamflow time series prediction , 2010, Eng. Appl. Artif. Intell..

[2]  Daniel Rueckert,et al.  Convolutional Recurrent Neural Networks for Dynamic MR Image Reconstruction , 2017, IEEE Transactions on Medical Imaging.

[3]  Lin Sun,et al.  Human Action Recognition Using Factorized Spatio-Temporal Convolutional Networks , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[4]  Kuldip K. Paliwal,et al.  Bidirectional recurrent neural networks , 1997, IEEE Trans. Signal Process..

[5]  Yan Liu,et al.  Recurrent Neural Networks for Multivariate Time Series with Missing Values , 2016, Scientific Reports.

[6]  J. Masiero,et al.  ASTEROID FAMILY IDENTIFICATION USING THE HIERARCHICAL CLUSTERING METHOD AND WISE/NEOWISE PHYSICAL PROPERTIES , 2013, 1305.1607.

[7]  Hongguang Li,et al.  Furnace thermal efficiency modeling using an improved convolution neural network based on parameter-adaptive mnemonic enhancement optimization , 2019, Applied Thermal Engineering.

[8]  Wei Shen,et al.  Spatial-temporal convolutional neural networks for anomaly detection and localization in crowded scenes , 2016, Signal Process. Image Commun..

[9]  Hans-Peter Kriegel,et al.  Density‐based clustering , 2011, WIREs Data Mining Knowl. Discov..

[10]  Krishna R. Pattipati,et al.  Data-Driven Modeling, Fault Diagnosis and Optimal Sensor Selection for HVAC Chillers , 2007, IEEE Transactions on Automation Science and Engineering.

[11]  Sullivan Hidot,et al.  An Expectation-Maximization algorithm for the Wishart mixture model: Application to movement clustering , 2010, Pattern Recognit. Lett..

[12]  Yoshua Bengio,et al.  Generative Adversarial Nets , 2014, NIPS.

[13]  Fionn Murtagh,et al.  Ward’s Hierarchical Agglomerative Clustering Method: Which Algorithms Implement Ward’s Criterion? , 2011, Journal of Classification.

[14]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[15]  Traian Iliescu,et al.  Data-Driven Filtered Reduced Order Modeling of Fluid Flows , 2017, SIAM J. Sci. Comput..

[16]  Joel Emer,et al.  Eyeriss: an Energy-efficient Reconfigurable Accelerator for Deep Convolutional Neural Networks Accessed Terms of Use , 2022 .

[17]  Ah Chung Tsoi,et al.  Face recognition: a convolutional neural-network approach , 1997, IEEE Trans. Neural Networks.

[18]  Yoshua Bengio,et al.  Understanding the difficulty of training deep feedforward neural networks , 2010, AISTATS.

[19]  S. C. Johnson.  Hierarchical clustering schemes , 1967, Psychometrika.

[20]  Nikolaos V. Sahinidis,et al.  A combined first-principles and data-driven approach to model building , 2015, Comput. Chem. Eng..

[21]  Gregory D. Hager,et al.  Temporal Convolutional Networks for Action Segmentation and Detection , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[22]  Nikos A. Vlassis,et al.  The global k-means clustering algorithm , 2003, Pattern Recognit..

[23]  Ping Zhang,et al.  A comparison study of basic data-driven fault diagnosis and process monitoring methods on the benchmark Tennessee Eastman process , 2012 .

[24]  Wang Qian,et al.  A graph-based clustering algorithm for anomaly intrusion detection , 2012, 2012 7th International Conference on Computer Science & Education (ICCSE).

[25]  Yongjian Wang,et al.  A novel intelligent modeling framework integrating convolutional neural network with an adaptive time-series window and its application to industrial process operational optimization , 2018, Chemometrics and Intelligent Laboratory Systems.

[26]  Hongguang Li,et al.  Complex Chemical Process Evaluation Methods Using a New Analytic Hierarchy Process Model Integrating Deep Residual Network with Multiway Principal Component Analysis , 2019 .