Improvement of Propeller Hydrodynamic Prediction Model Based on Multitask ANN and Its Application in Optimization Design
A multitask learning (MTL) model based on artificial neural networks (ANNs) is proposed in this study to improve the prediction accuracy and physical reliability of marine propeller hydrodynamic performance. The propeller’s comprehensive geometric features are used as inputs, and the coefficients of...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-01-01
|
| Series: | Journal of Marine Science and Engineering |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2077-1312/13/1/183 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832588202760732672 |
|---|---|
| author | Liang Li Yihong Chen Lu Huang Qing Hai Denghai Tang Chao Wang |
| author_facet | Liang Li Yihong Chen Lu Huang Qing Hai Denghai Tang Chao Wang |
| author_sort | Liang Li |
| collection | DOAJ |
| description | A multitask learning (MTL) model based on artificial neural networks (ANNs) is proposed in this study to improve the prediction accuracy and physical reliability of marine propeller hydrodynamic performance. The propeller’s comprehensive geometric features are used as inputs, and the coefficients of quadratic polynomials for the thrust coefficient (<i>K<sub>T</sub></i>) and torque coefficient (10<i>K<sub>Q</sub></i>) curves are predicted as outputs. The loss function is customized through a positive gradient penalty of the curves to accelerate training. When the single-task and multitask models were compared, the prediction errors were reduced; <i>K<sub>T</sub></i> decreased from 2.61% to 2.07%, 10 <i>K<sub>Q</sub></i> decreased from 3.58% to 2.31%, and the efficiency (<i>η</i>) decreased from 3.04% to 2.00%. Non-physical fluctuations in the performance curves were effectively mitigated by the multitask model, yielding predicted curvatures which closely matched the experimental data. Strong generalization was demonstrated when the model was tested on unseen propellers, with deviations of 2.2% for <i>K<sub>T</sub></i>, 4.6% for 10 <i>K<sub>Q</sub></i>, and 3.8% for <i>η</i>. Finally, the model was applied to optimize the propeller design for a 325,000 ton very large ore carrier ship, where a Pareto front with 58 non-dominant solutions for the maximum speed and fluctuating pressure was successfully generated and effectively verified by the model’s test results. The model enhanced the prediction of the propeller performance and contributed to optimization in the propeller’s design. |
| format | Article |
| id | doaj-art-639d9077ebb94dbe941b507c0f649086 |
| institution | Kabale University |
| issn | 2077-1312 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Journal of Marine Science and Engineering |
| spelling | doaj-art-639d9077ebb94dbe941b507c0f6490862025-01-24T13:37:10ZengMDPI AGJournal of Marine Science and Engineering2077-13122025-01-0113118310.3390/jmse13010183Improvement of Propeller Hydrodynamic Prediction Model Based on Multitask ANN and Its Application in Optimization DesignLiang Li0Yihong Chen1Lu Huang2Qing Hai3Denghai Tang4Chao Wang5College of Shipbuilding Engineering, Harbin Engineering University, Harbin 150001, ChinaChina Ship Scientific Research Center, Wuxi 214082, ChinaChina Ship Scientific Research Center, Wuxi 214082, ChinaChina Ship Scientific Research Center, Wuxi 214082, ChinaChina Ship Scientific Research Center, Wuxi 214082, ChinaCollege of Shipbuilding Engineering, Harbin Engineering University, Harbin 150001, ChinaA multitask learning (MTL) model based on artificial neural networks (ANNs) is proposed in this study to improve the prediction accuracy and physical reliability of marine propeller hydrodynamic performance. The propeller’s comprehensive geometric features are used as inputs, and the coefficients of quadratic polynomials for the thrust coefficient (<i>K<sub>T</sub></i>) and torque coefficient (10<i>K<sub>Q</sub></i>) curves are predicted as outputs. The loss function is customized through a positive gradient penalty of the curves to accelerate training. When the single-task and multitask models were compared, the prediction errors were reduced; <i>K<sub>T</sub></i> decreased from 2.61% to 2.07%, 10 <i>K<sub>Q</sub></i> decreased from 3.58% to 2.31%, and the efficiency (<i>η</i>) decreased from 3.04% to 2.00%. Non-physical fluctuations in the performance curves were effectively mitigated by the multitask model, yielding predicted curvatures which closely matched the experimental data. Strong generalization was demonstrated when the model was tested on unseen propellers, with deviations of 2.2% for <i>K<sub>T</sub></i>, 4.6% for 10 <i>K<sub>Q</sub></i>, and 3.8% for <i>η</i>. Finally, the model was applied to optimize the propeller design for a 325,000 ton very large ore carrier ship, where a Pareto front with 58 non-dominant solutions for the maximum speed and fluctuating pressure was successfully generated and effectively verified by the model’s test results. The model enhanced the prediction of the propeller performance and contributed to optimization in the propeller’s design.https://www.mdpi.com/2077-1312/13/1/183multitask learningfast predictionmarine propelleroptimization designopen-water performance |
| spellingShingle | Liang Li Yihong Chen Lu Huang Qing Hai Denghai Tang Chao Wang Improvement of Propeller Hydrodynamic Prediction Model Based on Multitask ANN and Its Application in Optimization Design Journal of Marine Science and Engineering multitask learning fast prediction marine propeller optimization design open-water performance |
| title | Improvement of Propeller Hydrodynamic Prediction Model Based on Multitask ANN and Its Application in Optimization Design |
| title_full | Improvement of Propeller Hydrodynamic Prediction Model Based on Multitask ANN and Its Application in Optimization Design |
| title_fullStr | Improvement of Propeller Hydrodynamic Prediction Model Based on Multitask ANN and Its Application in Optimization Design |
| title_full_unstemmed | Improvement of Propeller Hydrodynamic Prediction Model Based on Multitask ANN and Its Application in Optimization Design |
| title_short | Improvement of Propeller Hydrodynamic Prediction Model Based on Multitask ANN and Its Application in Optimization Design |
| title_sort | improvement of propeller hydrodynamic prediction model based on multitask ann and its application in optimization design |
| topic | multitask learning fast prediction marine propeller optimization design open-water performance |
| url | https://www.mdpi.com/2077-1312/13/1/183 |
| work_keys_str_mv | AT liangli improvementofpropellerhydrodynamicpredictionmodelbasedonmultitaskannanditsapplicationinoptimizationdesign AT yihongchen improvementofpropellerhydrodynamicpredictionmodelbasedonmultitaskannanditsapplicationinoptimizationdesign AT luhuang improvementofpropellerhydrodynamicpredictionmodelbasedonmultitaskannanditsapplicationinoptimizationdesign AT qinghai improvementofpropellerhydrodynamicpredictionmodelbasedonmultitaskannanditsapplicationinoptimizationdesign AT denghaitang improvementofpropellerhydrodynamicpredictionmodelbasedonmultitaskannanditsapplicationinoptimizationdesign AT chaowang improvementofpropellerhydrodynamicpredictionmodelbasedonmultitaskannanditsapplicationinoptimizationdesign |