A Comprehensive Evaluation of Physics-Informed Long Short-Term Memory Model for Ocean Current Prediction
Date
2024-07-07
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Department of Mechanical and Production Engineering(MPE), Islamic University of Technology(IUT), Board Bazar, Gazipur-1704, Bangladesh
Abstract
In an era where renewable energy sources are essential for sustainable development and
ensuring the quality of life for future generations, ocean energy has emerged as a significant
research focus. This is due to its vast potential for energy production from currents, waves, and
tides. A primary challenge in harnessing this energy lies in the inadequacy of accurate
assessments using near-real-world ocean models. This study addresses this challenge through
the implementation of Physics-Informed Long Short-Term Memory (PI-LSTM) models, which
integrate physics-based equations with machine learning techniques. The dataset used in this
study comprises 25,600 data points, with 100 time points for each of the 256 location points,
ranging within [0,1) for time and [-1,1] for location. Simulation data was generated by solving
Burger’s 1D equation, and a custom loss function was incorporated to embed Burger’s equation
as a physics loss term to train the model. The study's primary objective was to predict the
velocity component and evaluate the performance of PI-LSTM models against traditional
LSTM models and current Physics-Informed Neural Network (PINN) models based on
validation loss values. Results demonstrated that PI-LSTM models could predict the velocity
component with a Mean Squared Error (MSE) as low as 6.7e-5, outperforming basic LSTM
models in 55.56% of cases and PINN models in 77.78% of cases. Additionally, PI-LSTM
models required only half the number of epochs in the best cases compared to basic LSTM
models, indicating superior training efficiency. The findings suggest that PI-LSTM models not
only achieve lower validation loss values but also demonstrate enhanced training efficiency by
incorporating physical laws through custom loss functions, capturing the underlying dynamics
of oceanic processes more accurately with reduced computational effort. In conclusion, PI LSTM models show considerable potential as a superior tool for modeling ocean energy
systems, offering improved predictive performance and lower computational costs. This study
underscores the viability of PI-LSTM models in advancing ocean energy research, providing a
robust framework for more effective energy assessment and harnessing, and highlighting the
promising impact of integrating physics-informed methodologies within machine learning
models for the future of renewable energy
Description
Supervised by
Dr. Mohammad Ahsan Habib,
Professor,
Department of Production and Mechanical Engineering(MPE),
Islamic University of Technology (IUT)
Board Bazar, Gazipur-1704, Bangladesh
This thesis is submitted in partial fulfillment of the requirements for the degree of Bachelor of Science in Mechanical Engineering, 2024
Keywords
Citation
[1] “Scaling up investments in ocean energy technologies.” Accessed: Jan. 16, 2024. [Online]. Available: https://www.irena.org/Publications/2023/Mar/Scaling-up investments-in-ocean-energy-technologies [2] “OES | What is Ocean Energy | Tidal & Currents.” Accessed: Jan. 16, 2024. [Online]. Available: https://www.ocean-energy-systems.org/ocean-energy/what-is-ocean energy/tidal-currents/ [3] NOAA, “What is a current?”, National Ocean Service website. Accessed: Jan. 16, 2024. [Online]. Available: https://oceanservice.noaa.gov/facts/current.html [4] NOAA, “How do we monitor currents?”, National Ocean Service website. Accessed: Jan. 16, 2024. [Online]. Available: https://oceanservice.noaa.gov/facts/currentmon.html [5] X. Yang, K. A. Haas, and H. M. Fritz, “Theoretical assessment of ocean current energy potential for the Gulf Stream system,” Mar Technol Soc J, vol. 47, no. 4, pp. 101–112, 2013, doi: 10.4031/MTSJ.47.4.3. [6] J. C. Alcérreca-Huerta et al., “Energy yield assessment from ocean currents in the insular shelf of Cozumel Island,” J Mar Sci Eng, vol. 7, no. 5, May 2019, doi: 10.3390/jmse7050147. [7] J. F. Bárcenas Graniel, J. V. H. Fontes, H. F. Gomez Garcia, and R. Silva, “Assessing hydrokinetic energy in the mexican caribbean: A case study in the cozumel channel,” Energies (Basel), vol. 14, no. 15, Aug. 2021, doi: 10.3390/en14154411. [8] T. Xu, K. A. Haas, and B. Gunawan, “Estimating annual energy production from short tidal current records,” Renew Energy, vol. 207, pp. 105–115, May 2023, doi: 10.1016/j.renene.2023.02.107. [9] N. Y. Yahya, Silvianita, and D. Satrio, “A Review of Technology Development and Numerical Model Equation for Ocean Current Energy,” in IOP Conference Series: Earth and Environmental Science, Institute of Physics, 2023. doi: 10.1088/1755- 1315/1198/1/012019. [10] A. Khoshrou, A. B. Dorsman, and E. J. Pauwels, “SVD-based Visualisation and Approximation for Time Series Data in Smart Energy Systems.” [11] E. Can, E. Erku¸s, and V. Purutçuo˘ Glu, “Outlier Detection and Quasi-periodicity Optimization Algorithm: Frequency Domain Based Outlier Detection (FOD).” [12] T. W. Joo and S. B. Kim, “Time series forecasting based on wavelet filtering,” Expert Syst Appl, vol. 42, no. 8, pp. 3868–3874, May 2015, doi: 10.1016/j.eswa.2015.01.026. [13] D. B. Haidvogel et al., “Ocean forecasting in terrain-following coordinates: Formulation and skill assessment of the Regional Ocean Modeling System,” J Comput Phys, vol. 227, no. 7, pp. 3595–3624, Mar. 2008, doi: 10.1016/j.jcp.2007.06.016. [14] J. H. Vanzwieten, I. Meyer, and G. M. Alsenas, “EVALUATION OF HYCOM AS A TOOL FOR OCEAN CURRENT ENERGY ASSESSMENT.” [Online]. Available: http://www.hycom.org. © Department of Mechanical and Production Engineering IUT CHAPTER 6: BIBLIOGRAPHY 56 [15] E. P. Chassignet et al., “Global ocean prediction with the hybrid Coordinate Ocean Model (HYCOM),” Oceanography, vol. 22, no. SPL.ISS. 2, pp. 64–75, Jun. 2009, doi: 10.5670/oceanog.2009.39. [16] R. Pongto et al., “The Grid-Based Spatial ARIMA Model: An Innovation for Short-Term Predictions of Ocean Current Patterns with Big HF Radar Data,” in Advances in Intelligent Systems and Computing, Springer Verlag, 2020, pp. 26–36. doi: 10.1007/978- 3-030-19861-9_3. [17] S. Dauji, M. C. Deo, and K. Bhargava, “Prediction of ocean currents with artificial neural networks,” ISH Journal of Hydraulic Engineering, vol. 21, no. 1, pp. 14–27, Jan. 2015, doi: 10.1080/09715010.2014.938133. [18] C. Bayindir, “Predicting the Ocean Currents using Deep Learning,” Jun. 2019, [Online]. Available: http://arxiv.org/abs/1906.08066 [19] N. Thongniran, P. Vateekul, K. Jitkajornwanich, S. Lawawirojwong, and P. Srestasathiern, “Spatio-Temporal Deep Learning for Ocean Current Prediction Based on HF Radar Data,” in 2019 16th International Joint Conference on Computer Science and Software Engineering (JCSSE), IEEE, Jul. 2019, pp. 254–259. doi: 10.1109/JCSSE.2019.8864215. [20] A. Immas, N. Do, and M. R. Alam, “Real-time in situ prediction of ocean currents,” Ocean Engineering, vol. 228, May 2021, doi: 10.1016/j.oceaneng.2021.108922. [21] D. Saha, M. C. Deo, S. Joseph, and K. Bhargava, “A combined numerical and neural technique for short term prediction of ocean currents in the Indian Ocean,” Environmental Systems Research, vol. 5, no. 1, Dec. 2016, doi: 10.1186/s40068-016- 0057-2. [22] T. de Wolff, H. Carrillo, L. Martí, and N. Sanchez-Pi, “Towards Optimally Weighted Physics-Informed Neural Networks in Ocean Modelling,” Jun. 2021, [Online]. Available: http://arxiv.org/abs/2106.08747 [23] M. C. A. Clare, M. Sonnewald, R. Lguensat, J. Deshayes, and V. Balaji, “Explainable Artificial Intelligence for Bayesian Neural Networks: Toward Trustworthy Predictions of Ocean Dynamics,” J Adv Model Earth Syst, vol. 14, no. 11, Nov. 2022, doi: 10.1029/2022MS003162. [24] J. L. Wang, H. Zhuang, L. M. Chérubin, A. K. Ibrahim, and A. Muhamed Ali, “Medium Term Forecasting of Loop Current Eddy Cameron and Eddy Darwin Formation in the Gulf of Mexico With a Divide-and-Conquer Machine Learning Approach,” J Geophys Res Oceans, vol. 124, no. 8, pp. 5586–5606, Aug. 2019, doi: 10.1029/2019JC015172. [25] G. Yang, H. Wang, H. Qian, and J. Fang, “Tidal Current Short-Term Prediction Based on Support Vector Regression,” in IOP Conference Series: Materials Science and Engineering, Institute of Physics Publishing, May 2017. doi: 10.1088/1757- 899X/199/1/012024. [26] A. Kavousi-Fard and W. Su, “A Combined Prognostic Model Based on Machine Learning for Tidal Current Prediction,” IEEE Transactions on Geoscience and Remote Sensing, vol. 55, no. 6, pp. 3108–3114, Jun. 2017, doi: 10.1109/TGRS.2017.2659538. © Department of Mechanical and Production Engineering IUT CHAPTER 6: BIBLIOGRAPHY 57 [27] D. Sarkar, M. A. Osborne, and T. A. A. Adcock, “Spatiotemporal Prediction of Tidal Currents Using Gaussian Processes,” J Geophys Res Oceans, vol. 124, no. 4, pp. 2697– 2715, Apr. 2019, doi: 10.1029/2018JC014471. [28] D. Sarkar, M. A. Osborne, and T. A. A. Adcock, “Prediction of tidal currents using Bayesian machine learning,” Ocean Engineering, vol. 158, pp. 221–231, Jun. 2018, doi: 10.1016/j.oceaneng.2018.03.007. [29] P. G. Remya, R. Kumar, and S. Basu, “Forecasting tidal currents from tidal levels using genetic algorithm,” Ocean Engineering, vol. 40, pp. 62–68, Feb. 2012, doi: 10.1016/j.oceaneng.2011.12.002. [30] NOAA, “CO-OPS Current Station Data,” National Ocean Service website. Accessed: Jan. 17, 2024. [Online]. Available: https://tidesandcurrents.noaa.gov/cdata/StationList?type=Current+Data&filter=historic [31] “Inria-Chile/assessing-pinns-ocean-modelling: Code and datasets for paper ‘Assessing Weighted Physics Informed Neural Networks in Ocean Modelling.’” Accessed: May 29, 2024. [Online]. Available: https://github.com/Inria-Chile/assessing-pinns-ocean modelling/tree/main?tab=readme-ov-file [32] A. Khoshrou, A. B. Dorsman, and Eric. J. Pauwels, “SVD-based Visualisation and Approximation for Time Series Data in Smart Energy Systems,” Jul. 2018, doi: 10.1109/ISGTEurope.2017.8260303. [33] A. Trembanis, M. Lundine, and K. McPherran, “Coastal Mapping and Monitoring,” in Encyclopedia of Geology, Elsevier, 2021, pp. 251–266. doi: 10.1016/B978-0-12- 409548-9.12466-2. [34] A. Immas, N. Do, and M. R. Alam, “Real-time in situ prediction of ocean currents,” Ocean Engineering, vol. 228, May 2021, doi: 10.1016/j.oceaneng.2021.108922. [35] Z. C. Lipton, J. Berkowitz, and C. Elkan, “A Critical Review of Recurrent Neural Networks for Sequence Learning,” May 2015, [Online]. Available: http://arxiv.org/abs/1506.00019