Probabilistic real-time deep-water natural gas hydrate dispersion modeling by using a novel hybrid deep learning approach

• Published in: Energy (Oxford) Vol. 219; p. 119572
• Main Authors: Shi, Jihao, Li, Junjie, Usmani, Asif Sohail, Zhu, Yuan, Chen, Guoming, Yang, Dongdong
• Format: Journal Article
• Language: English
• Published: Oxford Elsevier Ltd 15.03.2021; Elsevier BV
• Subjects: Bayesian analysis; Bayesian theory; Computational fluid dynamics; Computer applications; Computer networks; Convolution variational autoencoder; data collection; Deep learning; Deep water; Digital twin of emergency management; Dispersion; Emergency management; Emergency preparedness; Emergency response; energy; fluid mechanics; gas hydrate; Gas hydrates; hybrids; Marine natural hydrate gas; Mathematical models; Model accuracy; Natural gas; Natural gas exploration; Neural networks; Probabilistic dispersion modeling; Probability theory; Real time; Uncertainty; Uncertainty estimation of spatial features; Variational Bayesian neural network; Marine natural hydrate gas; Probabilistic dispersion modeling; Uncertainty estimation of spatial features; Convolution variational autoencoder; Variational Bayesian neural network; Digital twin of emergency management
• ISSN: 0360-5442; 1873-6785
• DOI: 10.1016/j.energy.2020.119572

Computational Fluid Dynamic (CFD) has been widely used for the gas release and dispersion modeling, which however could not support real-time emergency response planning due to its high computation overhead. Surrogate models offer a potential alternative to rigorous computational approaches, however, as the point-estimation alternatives, the existing neural network-based surrogate models are not able to quantify the uncertainty of the released gas spatial concentration. This study aims to fill a gap by proposing an advanced hybrid probabilistic Convolutional-Variational Autoencoder-Variational Bayesian neural network (Conv-VAE-VBnn). Experimental study based on a benchmark simulation dataset was conducted. The results demonstrated the additional uncertainty information estimated by the proposed model contributes to reducing the harmful effect of too ‘confidence’ of the point-estimation models. In addition, the proposed model exhibits competitive accuracy with R2 = 0.94 compared and real-time capacity with inference time less than 1s. Latent size Nz = 2, noise σz=0.1 and Monte Carlo sampling number m = 500 to ensure the model’s real-time capacity, were also given. Overall, our proposed model could provide a reliable alternative for constructing a digital twin for emergency management during the exploration and exploitation of marine natural gas hydrate (NHG) in the near future. •Advanced probabilistic hybrid Conv-VAE-VBnn model is proposed.•Model correlates points with distribution of high dimensional spatial features.•Model quantifies uncertainty of spatial concentration given scenario-related inputs.•Model exhibits competitive accuracy and superior real-time application capability.•Hyper-parameters influencing model real-time application capability are analyzed.