Inversion Method Based on Temporal Convolutional Networks for Random Ice Load on Conical Offshore Platforms
The randomness and complexity of ice loads present major challenges to the safety and stability of offshore platforms. Traditional methods for identifying ice loads often lack accuracy and adaptability under changing environmental conditions. This study proposes a novel inversion method based on Tem...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-05-01
|
| Series: | Journal of Marine Science and Engineering |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2077-1312/13/5/1000 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850257544437563392 |
|---|---|
| author | Wei Li Ya Guo Shuzhao Li Yang Gao Yan Qu |
| author_facet | Wei Li Ya Guo Shuzhao Li Yang Gao Yan Qu |
| author_sort | Wei Li |
| collection | DOAJ |
| description | The randomness and complexity of ice loads present major challenges to the safety and stability of offshore platforms. Traditional methods for identifying ice loads often lack accuracy and adaptability under changing environmental conditions. This study proposes a novel inversion method based on Temporal Convolutional Networks (TCNs), integrating finite element simulation with deep learning to effectively identify random ice loads. A random ice load model is first developed, and its dynamic characteristics are validated through finite element analysis. The TCN model is then applied to capture the time-dependent features of ice loads. To improve the model’s generalization ability, its hyperparameters are optimized using particle swarm optimization (PSO). The results show that the TCN model achieves goodness-of-fit (R<sup>2</sup>) values of 0.821 and 0.808 on the training and test sets, respectively, indicating strong predictive performance. Under different ice thickness and velocity conditions, the model achieves R<sup>2</sup> values close to 0.99, demonstrating high robustness. This work represents the first application of TCN to ice load identification. By combining it with simulation data, we offer a high-precision, data-driven approach for dynamic load identification, enhancing the efficiency and reliability of safety assessments for conical offshore platforms. |
| format | Article |
| id | doaj-art-01de3f2b7c554c29bb5e94bdb10e683d |
| institution | OA Journals |
| issn | 2077-1312 |
| language | English |
| publishDate | 2025-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Journal of Marine Science and Engineering |
| spelling | doaj-art-01de3f2b7c554c29bb5e94bdb10e683d2025-08-20T01:56:24ZengMDPI AGJournal of Marine Science and Engineering2077-13122025-05-01135100010.3390/jmse13051000Inversion Method Based on Temporal Convolutional Networks for Random Ice Load on Conical Offshore PlatformsWei Li0Ya Guo1Shuzhao Li2Yang Gao3Yan Qu4CNOOC Research Institute Co., Ltd., Beijing 100029, ChinaShip Standardization Center, China Institute of Marine Technology & Economy, Beijing 100081, ChinaCNOOC Research Institute Co., Ltd., Beijing 100029, ChinaCNOOC Research Institute Co., Ltd., Beijing 100029, ChinaSchool of Marin Science and Technology, South China University of Technology, Guangzhou 510640, ChinaThe randomness and complexity of ice loads present major challenges to the safety and stability of offshore platforms. Traditional methods for identifying ice loads often lack accuracy and adaptability under changing environmental conditions. This study proposes a novel inversion method based on Temporal Convolutional Networks (TCNs), integrating finite element simulation with deep learning to effectively identify random ice loads. A random ice load model is first developed, and its dynamic characteristics are validated through finite element analysis. The TCN model is then applied to capture the time-dependent features of ice loads. To improve the model’s generalization ability, its hyperparameters are optimized using particle swarm optimization (PSO). The results show that the TCN model achieves goodness-of-fit (R<sup>2</sup>) values of 0.821 and 0.808 on the training and test sets, respectively, indicating strong predictive performance. Under different ice thickness and velocity conditions, the model achieves R<sup>2</sup> values close to 0.99, demonstrating high robustness. This work represents the first application of TCN to ice load identification. By combining it with simulation data, we offer a high-precision, data-driven approach for dynamic load identification, enhancing the efficiency and reliability of safety assessments for conical offshore platforms.https://www.mdpi.com/2077-1312/13/5/1000ice load identificationtemporal convolutional networkrandom ice load modelfinite element simulationparticle swarm optimizationdeep learning |
| spellingShingle | Wei Li Ya Guo Shuzhao Li Yang Gao Yan Qu Inversion Method Based on Temporal Convolutional Networks for Random Ice Load on Conical Offshore Platforms Journal of Marine Science and Engineering ice load identification temporal convolutional network random ice load model finite element simulation particle swarm optimization deep learning |
| title | Inversion Method Based on Temporal Convolutional Networks for Random Ice Load on Conical Offshore Platforms |
| title_full | Inversion Method Based on Temporal Convolutional Networks for Random Ice Load on Conical Offshore Platforms |
| title_fullStr | Inversion Method Based on Temporal Convolutional Networks for Random Ice Load on Conical Offshore Platforms |
| title_full_unstemmed | Inversion Method Based on Temporal Convolutional Networks for Random Ice Load on Conical Offshore Platforms |
| title_short | Inversion Method Based on Temporal Convolutional Networks for Random Ice Load on Conical Offshore Platforms |
| title_sort | inversion method based on temporal convolutional networks for random ice load on conical offshore platforms |
| topic | ice load identification temporal convolutional network random ice load model finite element simulation particle swarm optimization deep learning |
| url | https://www.mdpi.com/2077-1312/13/5/1000 |
| work_keys_str_mv | AT weili inversionmethodbasedontemporalconvolutionalnetworksforrandomiceloadonconicaloffshoreplatforms AT yaguo inversionmethodbasedontemporalconvolutionalnetworksforrandomiceloadonconicaloffshoreplatforms AT shuzhaoli inversionmethodbasedontemporalconvolutionalnetworksforrandomiceloadonconicaloffshoreplatforms AT yanggao inversionmethodbasedontemporalconvolutionalnetworksforrandomiceloadonconicaloffshoreplatforms AT yanqu inversionmethodbasedontemporalconvolutionalnetworksforrandomiceloadonconicaloffshoreplatforms |