Neural network-based source tracking of chemical leaks with obstacles

doi:10.1016/j.cjche.2020.12.022

Abstract

Abstract: The leakage of hazardous gases poses a significant threat to public security and causes environmental damage. The effective and accurate source term estimation (STE) is necessary when a leakage accident occurs. However, most research generally assumes that no obstacles exist near the leak source, which is inappropriate in practical applications. To solve this problem, we propose two different frameworks to emphasize STE with obstacles based on artificial neural network (ANN) and convolutional neural network (CNN). Firstly, we build a CFD model to simulate the gas diffusion in obstacle scenarios and construct a benchmark dataset. Secondly, we define the structure of ANN by searching, then predict the concentration distribution of gas using the searched model, and optimize source term parameters by particle swarm optimization (PSO) with well-performed cost functions. Thirdly, we propose a one-step STE method based on CNN, which establishes a link between the concentration distribution and the location of obstacles. Finally, we propose a novel data processing method to process sensor data, which maps the concentration information into feature channels. The comprehensive experiments illustrate the performance and efficiency of the proposed methods.

Key words: Obstacle, Optimization, Neural networks, Feature extraction, Source term estimation, Computational fluid dynamics (CFD)

摘要： The leakage of hazardous gases poses a significant threat to public security and causes environmental damage. The effective and accurate source term estimation (STE) is necessary when a leakage accident occurs. However, most research generally assumes that no obstacles exist near the leak source, which is inappropriate in practical applications. To solve this problem, we propose two different frameworks to emphasize STE with obstacles based on artificial neural network (ANN) and convolutional neural network (CNN). Firstly, we build a CFD model to simulate the gas diffusion in obstacle scenarios and construct a benchmark dataset. Secondly, we define the structure of ANN by searching, then predict the concentration distribution of gas using the searched model, and optimize source term parameters by particle swarm optimization (PSO) with well-performed cost functions. Thirdly, we propose a one-step STE method based on CNN, which establishes a link between the concentration distribution and the location of obstacles. Finally, we propose a novel data processing method to process sensor data, which maps the concentration information into feature channels. The comprehensive experiments illustrate the performance and efficiency of the proposed methods.

关键词: Obstacle, Optimization, Neural networks, Feature extraction, Source term estimation, Computational fluid dynamics (CFD)

Qiaoyi Xu, Wenli Du, Jinjin Xu, Jikai Dong. Neural network-based source tracking of chemical leaks with obstacles[J]. Chinese Journal of Chemical Engineering, 2021, 33(5): 211-220.

Qiaoyi Xu, Wenli Du, Jinjin Xu, Jikai Dong. Neural network-based source tracking of chemical leaks with obstacles[J]. 中国化学工程学报, 2021, 33(5): 211-220.

References

[1] Y. Xu, Z. Wang, Q. Zhu, An improved hybrid genetic algorithm for chemical plant layout optimization with novel non-overlapping and toxic gas dispersion constraints, Chin. J. Chem. Eng. 21(2013) 412-419.
[2] C. Bosanquet, J. Pearson, The spread of smoke and gases from chimneys, Trans. Faraday Soc. 32(1936) 1249.
[3] H. Li, J. Zhang, J. Yi, Computational source term estimation of the Gaussian puff dispersion, Soft Comput. 23(2018) 59-75.
[4] M. Hutchinson, H. Oh, W.-H. Chen, A review of source term estimation methods for atmospheric dispersion events using static or mobile sensors, Inf. Fusion 36(2017) 130-148.
[5] F. Murena, Measuring air quality over large urban areas: Development and application of an air pollution index at the urban area of Naples, Atmos. Environ. 38(2004) 6195-6202.
[6] J.M. Stockie, The mathematics of atmospheric dispersion modeling, SIAM Rev. 53(2011) 349-372.
[7] X. Li, J. Du, L. Wang, J. Fan, X. Peng, Effects of different nozzle materials on atomization results via CFD simulation, Chin. J. Chem. Eng. 28(2019) 362-368.
[8] Z. Zhang, X. Bao, Research status on inflow turbulence generation method with large eddy simulation of CFD numerical wind tunnel, in: Proceedings of IOP Conference Series: Materials Science and Engineering, vol. 490, IOP Publishing, (2019) 032015.
[9] A.O. de Souza, A.M. Luiz, A.T.P. Neto, A.C.B. de Araujo, H.B. da Silva, S.K. da Silva, J.J.N. Alves, CFD predictions for hazardous area classification, Chin. J. Chem. Eng. 27(2019) 21-31.
[10] S. Ye, Q. Tang, Y. Wang, W. Fei, Structural optimization of a settler via CFD simulation in a mixer-settler, Chin. J. Chem. Eng. 28(2020) 995-1015.
[11] S.R. Hanna, O.R. Hansen, M. Ichard, D. Strimaitis, CFD model simulation of dispersion from chlorine railcar releases in industrial and urban areas, Atmos. Environ. 43(2009) 262-270.
[12] J. Xing, Z. Liu, P. Huang, C. Feng, Y. Zhou, D. Zhang, F. Wang, Experimental and numerical study of the dispersion of carbon dioxide plume, J. Hazard. Mater. 256-257(2013) 40-48.
[13] B. Wang, F. Qian, W. Zhong, Wind field reconstruction for the dispersion modeling of accidental chemical spills on complex geometry, Chin. J. Chem. Eng. 27(2019) 2712-2724.
[14] V.M. Krasnopolsky, H. Schiller, Some neural network applications in environmental sciences. Part I: Forward and inverse problems in geophysical remote measurements, Neural Netw. 16(2003) 321-334.
[15] P.E. Bieringer, G.S. Young, L.M. Rodriguez, A.J. Annunzio, F. Vandenberghe, S.E. Haupt, Paradigms and commonalities in atmospheric source term estimation methods, Atmos. Environ. 156(2017) 102-112.
[16] E. Soroush, S. Shahsavari, M. Mesbah, M. Rezakazemi, Z. Zhang, A robust predictive tool for estimating CO₂ solubility in potassium based amino acid salt solutions, Chin. J. Chem. Eng. 26(2018) 740-746.
[17] S. Reich, D. Gomez, L. Dawidowski, Artificial neural network for the identification of unknown air pollution sources, Atmos. Environ. 33(1999) 3045-3052.
[18] D. Ma, Z. Zhang, Contaminant dispersion prediction and source estimation with integrated Gaussian-machine learning network model for point source emission in atmosphere, J. Hazard. Mater. 311(2016) 237-245.
[19] Y. Shi, R. Eberhart, Empirical study of particle swarm optimization, in: Proceedings of the 1999 congress on evolutionary computation-CEC99, vol. 3, IEEE, 1999, pp. 1945-1950.
[20] Y. Shi, R. Eberhart, Particle swarm optimization: Developments, applications and resources, in: Proceedings of the 2001 congress on evolutionary computation, vol. 1, IEEE, (2001) 81-86.
[21] D. Ma, W. Tan, Z. Zhang, J. Hu, Gas emission source term estimation with 1-step nonlinear partial swarm optimization-tikhonov regularization hybrid method, Chin. J. Chem. Eng. 26(2018) 356-363.
[22] S. Qiu, B. Chen, R. Wang, Z. Zhu, Y. Wang, X. Qiu, Atmospheric dispersion prediction and source estimation of hazardous gas using artificial neural network, particle swarm optimization and expectation maximization, Atmos. Environ. 178(2018) 158-163.
[23] T. Moon, The expectation-maximization algorithm, IEEE Signal Process. Mag. 13(1996) 47-60.
[24] Y. Wang, H. Huang, L. Huang, X. Zhang, Source term estimation of hazardous material releases using hybrid genetic algorithm with composite cost functions, Eng. Appl. Artif. Intell. 75(2018) 102-113.
[25] Y. LeCun, Y. Bengio, G. Hinton, Deep learning, Nature 521(2015) 436-444.
[26] D. Silver, A. Huang, C.J. Maddison, A. Guez, L. Sifre, G. van den Driessche, J. Schrittwieser, I. Antonoglou, V. Panneershelvam, M. Lanctot, S. Dieleman, D. Grewe, J. Nham, N. Kalchbrenner, I. Sutskever, T. Lillicrap, M. Leach, K. Kavukcuoglu, T. Graepel, D. Hassabis, Mastering the game of go with deep neural networks and tree search, Nature 529(2016) 484-489.
[27] Y. Jin, H. Wang, T. Chugh, D. Guo, K. Miettinen, Data-driven evolutionary optimization: An overview and case studies, IEEE Trans. Evol. Computat. 23(2019) 442-458.
[28] X. Chen, W. Du, F. Qian, Solving chemical dynamic optimization problems with ranking-based differential evolution algorithms, Chin. J. Chem. Eng. 24(2016) 1600-1608.
[29] F. Sun, W. Du, R. Qi, F. Qian, W. Zhong, A hybrid improved genetic algorithm and its application in dynamic optimization problems of chemical processes, Chin. J. Chem. Eng. 21(2013) 144-154.
[30] X. Glorot, Y. Bengio, Understanding the difficulty of training deep feedforward neural networks, in: Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, Sardinia, Italy, 2010, 249-256.
[31] C. Li, D. Zhao, S. Mu, W. Zhang, N. Shi, L. Li, Fault diagnosis for distillation process based on CNN-DAE, Chin. J. Chem. Eng. 27(2019) 598-604.
[32] G. Li, Y. Yu, Visual saliency detection based on multiscale deep CNN features, IEEE Trans. Image Process. 25(2016) 5012-5024.
[33] J. Cho, H. Kim, A.L. Gebreselassie, D. Shin, Deep neural network and random forest classifier for source tracking of chemical leaks using fence monitoring data, J. Loss Prev. Process Ind. 56(2018) 548-558.
[34] B. Wang, F. Qian, Three dimensional gas dispersion modeling using cellular automata and artificial neural network in urban environment, Process Saf. Environ. Protect. 120(2018) 286-301.
[35] R. Yoshie, A. Mochida, Y. Tominaga, H. Kataoka, K. Harimoto, T. Nozu, T. Shirasawa, Cooperative project for CFD prediction of pedestrian wind environment in the architectural institute of japan, J. Wind Eng. Ind. Aerodyn. 95(2007) 1551-1578.
[36] Y. Mouilleau, A. Champassith, CFD simulations of atmospheric gas dispersion using the fire dynamics simulator (FDS), J. Loss Prev. Process Ind. 22(2009) 316-323.
[37] K. Hornik, Approximation capabilities of multilayer feedforward networks, Neural Netw. 4(1991) 251-257.
[38] J. Kennedy, R. Eberhart, Particle swarm optimization, In: Proceedings of ICNN’95-International Conference on Neural Networks, vol. 4, Citeseer, IEEE, (1995) 1942-1948.
[39] A. Krizhevsky, I. Sutskever, G.E. Hinton, Imagenet classification with deep convolutional neural networks, Commun. ACM 60(2017) 84-90.
[40] J. Masci, U. Meier, D. Ciresan, J. Schmidhuber, Stacked convolutional autoencoders for hierarchical feature extraction, In: Proceedings of International Conference on Artificial Neural Networks, Lecture Notes in Computer Science Belin, Heidelberg: Springer Berlin Heidelbery, (2011) 52-59.
[41] D. Zhao, T. Wang, F. Chu, Deep convolutional neural network based planet bearing fault classification, Comput. Ind. 107(2019) 59-66.
[42] K. He, X. Zhang, S. Ren, J. Sun, Deep residual learning for image recognition, in: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, NV, USA, IEEE, (2016) 770-778.
[43] H. Wang, Y. Jin, C. Sun, J. Doherty, Offline data-driven evolutionary optimization using selective surrogate ensembles, IEEE Trans. Evol. Comput. 23(2019) 203-216.
[44] MATLAB, version 7.10.0(R2010a), The MathWorks Inc., Natick, Massachusetts, 2010.
[45] A. Paszke, S. Gross, S. Chintala, G. Chanan, E. Yang, Z. DeVito, Z. Lin, A. Desmaison, L. Antiga, A. Lerer, Automatic differentiation in pytorch, In: Proceedings of Advances in Neural Information Processing Systems, Long Beach, CA, USA, 2017.