Construction High Precision Neural Network Proxy Model for Ship Hull Structure Design Based on Hybrid Datasets of Hydrodynamic Loads

Abstract

In this work, we constructed a neural network proxy model (NNPM) to estimate the hydrodynamic resistance in the ship hull structure design process, which is based on the hydrodynamic load data obtained from both the potential flow method (PFM) and the viscous flow method (VFM). Here the PFM dataset is applied for the tuning, pre-training, and the VFM dataset is applied for the fine-training. By adopting the PFM and VFM datasets simultaneously, we aim to construct an NNPM to achieve the high-accuracy prediction on hydrodynamic load on ship hull structures exerted from the viscous flow, while ensuring a moderate data-acquiring workload. The high accuracy prediction on hydrodynamic loads and the relatively low dataset establishment cost of the NNPM developed demonstrated the effectiveness and feasibility of hybrid dataset based NNPM achieving a high precision prediction of hydrodynamic loads on ship hull structures. The successful construction of the high precision hydrodynamic prediction NNPM advances the artificial intelligence-assisted design (AIAD) technology for various marine structures.

Keywords

Deep learning neural network / Hybrid dataset / Proxy model / Ship hull design / Machine learning

References

[1] Alfonsi G. Reynolds-averaged Navier-Stokes equations for turbulence modeling. Applied Mechanics Reviews, 2009, 62(4): 040802

[2] Anderson JD, Wendt J (1995) Computational fluid dynamics. Volume 206 Springer

[3] Ao Y, Li Y, Gong J, Li S. An artificial intelligence-aided design (AIAD) of ship hull structures. Journal of Ocean Engineering and Science, 2021, 8(1): 15-32

[4] Ao Y, Li Y, Gong J, Li S. Artificial intelligence design for ship structures: A variant multiple-input neural network based ship resistance prediction. Journal of Mechanical Design, 2022, 144(9): 1-18

[5] Bakhtiari M, Ghassemi H. CFD data based neural network functions for predicting hydrodynamic performance of a low-pitch marine cycloidal propeller. Applied Ocean Research, 2020, 94: 101981

[6] Buhmann MD (2000). Radial basis functions. Acta Numerica 9: 1–38 Caruana R, Niculescu-Mizil A (2006) An empirical comparison of supervised learning algorithms. Proceedings of the 23rd International Conference on Machine Learning, 161–168

[7] Chorin AJ. A numerical method for solving incompressible viscous flow problems. Journal of Computational Physics, 1997, 135: 118-125

[8] Claveria O, Monte E, Torra S (2015) Multiple-input multiple-output vs. single-input single-output neural network forecasting. In: Research Institute of Applied Economics. Barcelona University, 2015-02

[9] Dawson C (1977) A practical computer method for solving ship-wave problems. Proceedings of Second International Conference on Numerical Ship Hydrodynamics, 30–38

[10] Dick S (2019) Artificial intelligence. Harvard Data Science Review 1.1

[11] Dodge J, Ilharco G, Schwartz R, Farhadi A, Hajishirzi H, Smith N (2020) Fine-tuning pretrained language models: Weight initializations, data orders, and early stopping. arXiv preprint arXiv:2002.06305

[12] Forti D, Rozza G. Efficient geometrical parametrisation techniques of interfaces for reduced-order modelling: Application to fluid-structure interaction coupling problems. International Journal of Computational Fluid Dynamics, 2014, 28: 158-169

[13] Goodfellow I, Bengio Y, Courville A (2016) Deep learning. MIT Press He K, Zhang X, Ren S, Sun J (2015) Delving deep into rectifiers: Surpassing human-level performance on imagenet classification. Proceedings of the IEEE International Conference on Computer Vision, 1026–1034

[14] Hess JL, Smith A. Calculation of potential flow about arbitrary bodies. Progress in Aerospace Sciences, 1967, 8: 1-138

[15] Huang CJ, Kuo PH. Multiple-input deep convolutional neural network model for short-term photovoltaic power forecasting. IEEE Access, 2019, 7: 74822-74834

[16] Kazemi H, Doustdar MM, Najafi A, Nowruzi H, Ameri MJ. Hydrodynamic performance prediction of stepped planing craft using CFD and ANNs. Journal of Marine Science and Application, 2021, 20: 67-84

[17] Kingma DP, Ba J (2014) Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980

[18] Li Y, Gong J, Ma Q, Yan S. Effects of the terms associated with φzz in free surface condition on the attitudes and resistance of different ships. Engineering Analysis with Boundary Elements, 2018, 95: 266-285

[19] Mittendorf M, Nielsen UD, Bingham HB. Data-driven prediction of added-wave resistance on ships in oblique waves comparison between tree-based ensemble methods and artificial neural networks. Applied Ocean Research, 2022, 118: 102964

[20] Nagelkerke NJ. A note on a general definition of the coefficient of determination. Biometrika, 1991, 78: 691-692

[21] Peng H, Ni S, Qiu W. Wave pattern and resistance prediction for ships of full form. Ocean Engineering, 2014, 87: 162-173

[22] Piegl L, Tiller W (1996) The NURBS book. Springer Science & Business Media

[23] Prpić-Oršić J, Valčić M, Čarija Z. A hybrid wind load estimation method for container ship based on computational fluid dynamics and neural networks. Journal of Marine Science and Engineering, 2020, 8: 539

[24] Schulz E, Speekenbrink M, Krause A. A tutorial on gaussian process regression: Modelling, exploring, and exploiting functions. Journal of Mathematical Psychology, 2018, 85: 1-16

[25] Services PE (2021) Siemens Digital Industries Software

[26] Shora MM, Ghassemi H, Nowruzi H. Using computational fluid dynamic and artificial neural networks to predict the performance and cavitation volume of a propeller under different geometrical and physical characteristics. Journal of Marine Engineering & Technology, 2018, 17: 59-84

[27] Silva KM, Maki KJ. Data-driven system identification of 6-DOF ship motion in waves with neural networks. Applied Ocean Research, 2022, 125: 103222

[28] Snoek J, Larochelle H, Adams RP. Practical Bayesian optimization of machine learning algorithms. Advances in Neural Information Processing Systems, 2012, 25: 2960-2968

[29] Specht DF. A general regression neural network. IEEE Transactions on Neural Networks, 1991, 2: 568-576

[30] Srivastava N, Hinton G, Krizhevsky A, Sutskever I, Salakhutdinov R. Dropout: a simple way to prevent neural networks from overfitting. The Journal of Machine Learning Research, 2014, 15: 1929-1958

[31] Zhang RZ. Verification and validation for RANS simulation of KCS container ship without/with propeller. Journal of Hydrodynamics, 2010, 22: 889-896