Ocean Engineering最新文献_第2页

Spatiotemporal synchronization-aware cross-domain mission planning for air-sea-underwater heterogeneous unmanned swarms 空-海-水异构无人蜂群的时空同步感知跨域任务规划

IF 5.5 2区工程技术 Q1 ENGINEERING, CIVIL

Ocean Engineering

Pub Date : 2026-04-30 Epub Date: 2026-02-20 DOI: 10.1016/j.oceaneng.2026.124694

Yishuo Meng , Songyan Xu , Xiaolong Liang , Shilin Zeng , Zhiwen Yao , Chengsi Bian , Lei Wang , Changting Zhong , Xinwei Wang

Facing the increasing complexity of multi-dimensional maritime operations, cross-domain heterogeneous unmanned swarms provide an appealing paradigm for comprehensive ocean perception. However, platform disparities and strict spatiotemporal synchronization requirements challenge mission planning where task assignment and path planning are tightly coupled. To address this, we design a single-stage cooperative planning framework tailored for air-sea-underwater swarms. First, a recursive temporal deduction mechanism is introduced to accurately quantify the cooperative waiting costs induced by platform heterogeneity. Subsequently, a Q-learning enhanced Genetic Algorithm integrating chaotic initialization and Opposition-Based Learning, named COGAQ, is proposed for rapid solution. This algorithm adopts a feasible coalition set index encoding strategy to limit the search space and integrates variable neighborhood search to rectify temporal asynchrony. Comparative experiments against mainstream algorithms validate the effectiveness of the designed framework, highlighting the optimum-seeking capability of COGAQ in handling tightly coupled constraints. Furthermore, a large-scale scenario containing 55 heterogeneous platforms and 102 targets verifies the algorithm's scalability. This paper provides a precise and efficient solution for air-sea-underwater heterogeneous temporal coordination problems, presenting an appealing approach for cooperative mission planning in complex scenarios like maritime search and rescue, ocean resource monitoring, and joint anti-submarine operations.

面对日益复杂的多维海上作战，跨域异构无人群为全面的海洋感知提供了一种极具吸引力的范式。然而，平台差异和严格的时空同步要求给任务分配和路径规划紧密耦合的任务规划带来了挑战。为了解决这个问题，我们设计了一个为空气-海洋-水下群体量身定制的单阶段合作规划框架。首先，引入递归时间演绎机制，准确量化平台异质性导致的合作等待成本；随后，为了快速求解该问题，提出了一种融合混沌初始化和基于对立学习的q学习增强遗传算法COGAQ。该算法采用可行的联盟集索引编码策略来限制搜索空间，并结合可变邻域搜索来纠正时间异步性。与主流算法的对比实验验证了所设计框架的有效性，突出了COGAQ在处理紧耦合约束时的寻优能力。此外，在包含55个异构平台和102个目标的大规模场景中验证了算法的可扩展性。本文提供了一种精确、高效的空-海-水下异构时间协调问题解决方案，为海上搜救、海洋资源监测、联合反潜作战等复杂场景下的协同任务规划提供了一种有吸引力的方法。

{"title":"Spatiotemporal synchronization-aware cross-domain mission planning for air-sea-underwater heterogeneous unmanned swarms","authors":"Yishuo Meng , Songyan Xu , Xiaolong Liang , Shilin Zeng , Zhiwen Yao , Chengsi Bian , Lei Wang , Changting Zhong , Xinwei Wang","doi":"10.1016/j.oceaneng.2026.124694","DOIUrl":"10.1016/j.oceaneng.2026.124694","url":null,"abstract":"<div><div>Facing the increasing complexity of multi-dimensional maritime operations, cross-domain heterogeneous unmanned swarms provide an appealing paradigm for comprehensive ocean perception. However, platform disparities and strict spatiotemporal synchronization requirements challenge mission planning where task assignment and path planning are tightly coupled. To address this, we design a single-stage cooperative planning framework tailored for air-sea-underwater swarms. First, a recursive temporal deduction mechanism is introduced to accurately quantify the cooperative waiting costs induced by platform heterogeneity. Subsequently, a Q-learning enhanced Genetic Algorithm integrating chaotic initialization and Opposition-Based Learning, named COGAQ, is proposed for rapid solution. This algorithm adopts a feasible coalition set index encoding strategy to limit the search space and integrates variable neighborhood search to rectify temporal asynchrony. Comparative experiments against mainstream algorithms validate the effectiveness of the designed framework, highlighting the optimum-seeking capability of COGAQ in handling tightly coupled constraints. Furthermore, a large-scale scenario containing 55 heterogeneous platforms and 102 targets verifies the algorithm's scalability. This paper provides a precise and efficient solution for air-sea-underwater heterogeneous temporal coordination problems, presenting an appealing approach for cooperative mission planning in complex scenarios like maritime search and rescue, ocean resource monitoring, and joint anti-submarine operations.</div></div>","PeriodicalId":19403,"journal":{"name":"Ocean Engineering","volume":"353 ","pages":"Article 124694"},"PeriodicalIF":5.5,"publicationDate":"2026-04-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147387259","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Numerical investigation of submarine surface scattering and scaling effects under point forces excitation 点力激励下潜艇表面散射和尺度效应的数值研究

IF 5.5 2区工程技术 Q1 ENGINEERING, CIVIL

Ocean Engineering

Pub Date : 2026-04-30 Epub Date: 2026-02-18 DOI: 10.1016/j.oceaneng.2026.124740

Xiaoshen Ning , Xiaorui Bai , Xusheng Li , Lin Li , Jianxin Ren , Qing Wang , Jian Hu

Underwater radiated noise (URN) is a critical factor in submarine stealth performance, with hydrodynamic loading acting as a major excitation source during high-speed cruising. This study investigates the role of non-compact hull surface scattering in shaping acoustic directivity patterns using a frequency-domain boundary element method (BEM). The method is validated against analytical solutions and applied to a DARPA Suboff subjected to multiple dipole excitations at the stern. Simulations were conducted across a range of Helmholtz numbers to assess frequency-dependent spatial directivity, and scaling analyses were performed to examine geometric similarity conditions. Results reveal that at low frequencies, directivity resembles the classical dipole figure-eight pattern, but when the wavelength becomes comparable to the hull dimensions, the non-compact boundary induces significant side lobes and compresses the main radiation lobe. Finally, these scattering effects persist into the far field in multi-scale configurations. The study confirms that spatial scale similarity can be preserved when source strength is scaled with the square of the geometric factor, but fixed-frequency evaluations expose scale-induced distortions in acoustic fields. These findings underscore the necessity of incorporating non-compact surface scattering into URN predictions, as neglecting such effects may result in inaccurate assessments of noise directivity in full-scale stealth evaluations.

水下辐射噪声（URN）是影响潜艇隐身性能的关键因素，而水下动力载荷是潜艇高速巡航时的主要激励源。本研究利用频域边界元法（BEM）研究非致密船体表面散射在形成声指向性模式中的作用。用解析解对该方法进行了验证，并应用于DARPA Suboff飞机艉部受到多重偶极子激励的情况。在一系列亥姆霍兹数范围内进行了模拟，以评估频率相关的空间指向性，并进行了缩放分析，以检查几何相似性条件。结果表明，在低频时，指向性类似于经典的偶极子八位数模式，但当波长与船体尺寸相当时，非紧致边界产生显著的侧瓣并压缩主辐射瓣。最后，这些散射效应在多尺度构型下持续到远场。研究证实，当声源强度按几何因子的平方进行缩放时，可以保持空间尺度的相似性，但固定频率的评估暴露了声场中由尺度引起的扭曲。这些发现强调了将非致密表面散射纳入URN预测的必要性，因为忽略这种影响可能导致在全面隐身评估中对噪声指向性的评估不准确。

{"title":"Numerical investigation of submarine surface scattering and scaling effects under point forces excitation","authors":"Xiaoshen Ning , Xiaorui Bai , Xusheng Li , Lin Li , Jianxin Ren , Qing Wang , Jian Hu","doi":"10.1016/j.oceaneng.2026.124740","DOIUrl":"10.1016/j.oceaneng.2026.124740","url":null,"abstract":"<div><div>Underwater radiated noise (URN) is a critical factor in submarine stealth performance, with hydrodynamic loading acting as a major excitation source during high-speed cruising. This study investigates the role of non-compact hull surface scattering in shaping acoustic directivity patterns using a frequency-domain boundary element method (BEM). The method is validated against analytical solutions and applied to a DARPA Suboff subjected to multiple dipole excitations at the stern. Simulations were conducted across a range of Helmholtz numbers to assess frequency-dependent spatial directivity, and scaling analyses were performed to examine geometric similarity conditions. Results reveal that at low frequencies, directivity resembles the classical dipole figure-eight pattern, but when the wavelength becomes comparable to the hull dimensions, the non-compact boundary induces significant side lobes and compresses the main radiation lobe. Finally, these scattering effects persist into the far field in multi-scale configurations. The study confirms that spatial scale similarity can be preserved when source strength is scaled with the square of the geometric factor, but fixed-frequency evaluations expose scale-induced distortions in acoustic fields. These findings underscore the necessity of incorporating non-compact surface scattering into URN predictions, as neglecting such effects may result in inaccurate assessments of noise directivity in full-scale stealth evaluations.</div></div>","PeriodicalId":19403,"journal":{"name":"Ocean Engineering","volume":"353 ","pages":"Article 124740"},"PeriodicalIF":5.5,"publicationDate":"2026-04-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147387095","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Assessing human reliability in reactive cargo handling for chemical tanker ships 化学品船反应性货物装卸人员可靠性评估

IF 5.5 2区工程技术 Q1 ENGINEERING, CIVIL

Ocean Engineering

Pub Date : 2026-04-30 Epub Date: 2026-02-23 DOI: 10.1016/j.oceaneng.2026.124780

Esma Uflaz , Muhittin Orhan , M. Fatih Gulen , Furkan Gumus , Ozcan Arslan

This study assesses human reliability in reactive cargo handling on chemical tankers and proposes a hybrid Dempster-Shafer (DS)-extended Success Likelihood Index Methodology (SLIM) approach to enhance process safety. A hierarchical task analysis decomposes the operation, DS theory aggregates expert data to address subjectivity and uncertainty, and SLIM derives task/subtask HEPs. Human error probabilities are aggregated at the system level to assess overall system reliability. The novelty lies in assessing an under-researched high-risk shipboard operation with a novel hybrid methodology and offering practical risk mitigation strategies. Key findings show task complexity as the most influential PSF, followed by training/experience. Inhibitor quantity calculations, continuous monitoring procedures, and thermal management operations are revealed as the most critical tasks. Consequently, the overall human reliability is calculated as 8.20E-01. The findings provide valuable insights for maritime stakeholders, suggesting targeted interventions in calculation protocols, monitoring systems, heat management strategies, and specialised crew training to enhance process safety in handling reactive cargo onboard chemical tankers.

本研究评估了化学品罐车反应性货物装卸过程中人类的可靠性，并提出了一种混合的Dempster-Shafer （DS）扩展成功可能性指数方法（SLIM）来提高过程安全性。分层任务分析分解操作，DS理论聚合专家数据以解决主观性和不确定性，SLIM理论派生任务/子任务hep。在系统级别汇总人为错误概率，以评估整体系统可靠性。其新颖之处在于用一种新颖的混合方法来评估研究不足的高风险船上作业，并提供实用的风险缓解策略。主要调查结果显示，任务复杂性是最具影响力的PSF，其次是培训/经验。缓蚀剂用量计算、连续监测程序和热管理操作是最关键的任务。因此，总体人的可靠性计算为8.20E-01。研究结果为海事利益相关者提供了宝贵的见解，建议在计算协议、监测系统、热管理策略和专业船员培训方面进行有针对性的干预，以提高化学品船上处理反应性货物的过程安全性。

{"title":"Assessing human reliability in reactive cargo handling for chemical tanker ships","authors":"Esma Uflaz , Muhittin Orhan , M. Fatih Gulen , Furkan Gumus , Ozcan Arslan","doi":"10.1016/j.oceaneng.2026.124780","DOIUrl":"10.1016/j.oceaneng.2026.124780","url":null,"abstract":"<div><div>This study assesses human reliability in reactive cargo handling on chemical tankers and proposes a hybrid Dempster-Shafer (DS)-extended Success Likelihood Index Methodology (SLIM) approach to enhance process safety. A hierarchical task analysis decomposes the operation, DS theory aggregates expert data to address subjectivity and uncertainty, and SLIM derives task/subtask HEPs. Human error probabilities are aggregated at the system level to assess overall system reliability. The novelty lies in assessing an under-researched high-risk shipboard operation with a novel hybrid methodology and offering practical risk mitigation strategies. Key findings show task complexity as the most influential PSF, followed by training/experience. Inhibitor quantity calculations, continuous monitoring procedures, and thermal management operations are revealed as the most critical tasks. Consequently, the overall human reliability is calculated as 8.20E-01. The findings provide valuable insights for maritime stakeholders, suggesting targeted interventions in calculation protocols, monitoring systems, heat management strategies, and specialised crew training to enhance process safety in handling reactive cargo onboard chemical tankers.</div></div>","PeriodicalId":19403,"journal":{"name":"Ocean Engineering","volume":"353 ","pages":"Article 124780"},"PeriodicalIF":5.5,"publicationDate":"2026-04-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147386611","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Fusing SWAN priors with tucker decomposition: A training-free framework for sparse wave field reconstruction 融合SWAN先验与tucker分解：稀疏波场重建的无训练框架

IF 5.5 2区工程技术 Q1 ENGINEERING, CIVIL

Ocean Engineering

Pub Date : 2026-04-30 Epub Date: 2026-02-21 DOI: 10.1016/j.oceaneng.2026.124786

Shanxun Sun , Zhicong Yang , Zihan Cai , Lianfang Ye , Yanni Kou , Ting He

Reconstructing complete multi-parameter wave fields from sparse observations is essential for marine energy applications and engineering safety. This study presents a training-free, physics-constrained approach that integrates SWAN-informed historical priors with Tucker low-rank decomposition, sparsity-based inversion, and temporal fusion to jointly reconstruct significant wave height, period, and direction. An objective PCA-weighted composite error metric is introduced to evaluate cross-variable performance. Using the South China Sea as the test area, multilinear ranks are optimized via Gaussian-process Bayesian search, and the sensor layout is designed to balance accuracy with cost-effectiveness. Using an optimized 140-sensor layout, month-long reconstructions in July 2023 achieved RMSE/MAE/R² of 0.1085/0.0609/0.9570 for Hsig (m), 0.2946/0.1686/0.9294 for Tm01 (s), and 0.1405/0.0705/0.9144 for Dir (rad). Under Typhoon Saola (2309), the method preserved robust directional reconstruction (RMSE = 0.1474 rad, R² = 0.7865) under rapid regime shifts. Compared with CNN, LSTM, TCN, and Transformer baselines in both swell-dominated and typhoon-dominated regimes, the proposed method yields lower composite errors and more stable performance, with particularly robust directional reconstruction enabled by circular encoding. These results demonstrate strong accuracy, stability, and adaptability under sparse deployments, providing practical insights for observing-system design and multi-source data fusion in ocean engineering.

从稀疏观测数据中重建完整的多参数波场对于海洋能源应用和工程安全至关重要。本研究提出了一种无需训练、物理约束的方法，该方法将swan信息的历史先验与Tucker低秩分解、基于稀疏性的反演和时间融合结合起来，共同重建重要的波高、周期和方向。引入了一种客观的pca加权复合误差度量来评价跨变量性能。以南海为试验区，通过高斯过程贝叶斯搜索优化多线性秩，设计传感器布局，兼顾精度和成本效益。使用优化的140个传感器布局，2023年7月的一个月重建结果显示，Hsig (m)的RMSE/MAE/R2为0.1085/0.0609/0.9570,Tm01 (s)的RMSE/MAE/R2为0.2946/0.1686/0.9294,Dir （rad）的RMSE/MAE/R2为0.1405/0.0705/0.9144。在台风“索拉”（2309）下，该方法在快速状态转移下保持了稳健的方向重建（RMSE = 0.1474 rad, R2 = 0.7865）。与CNN、LSTM、TCN和Transformer基线在膨胀主导和台风主导两种情况下相比，该方法产生更低的复合误差和更稳定的性能，并通过循环编码实现了特别鲁棒的方向重建。这些结果显示了稀疏部署下较强的准确性、稳定性和适应性，为海洋工程中观测系统设计和多源数据融合提供了实用的见解。

{"title":"Fusing SWAN priors with tucker decomposition: A training-free framework for sparse wave field reconstruction","authors":"Shanxun Sun , Zhicong Yang , Zihan Cai , Lianfang Ye , Yanni Kou , Ting He","doi":"10.1016/j.oceaneng.2026.124786","DOIUrl":"10.1016/j.oceaneng.2026.124786","url":null,"abstract":"<div><div>Reconstructing complete multi-parameter wave fields from sparse observations is essential for marine energy applications and engineering safety. This study presents a training-free, physics-constrained approach that integrates SWAN-informed historical priors with Tucker low-rank decomposition, sparsity-based inversion, and temporal fusion to jointly reconstruct significant wave height, period, and direction. An objective PCA-weighted composite error metric is introduced to evaluate cross-variable performance. Using the South China Sea as the test area, multilinear ranks are optimized via Gaussian-process Bayesian search, and the sensor layout is designed to balance accuracy with cost-effectiveness. Using an optimized 140-sensor layout, month-long reconstructions in July 2023 achieved RMSE/MAE/R<sup>2</sup> of 0.1085/0.0609/0.9570 for Hsig (m), 0.2946/0.1686/0.9294 for Tm01 (s), and 0.1405/0.0705/0.9144 for Dir (rad). Under Typhoon Saola (2309), the method preserved robust directional reconstruction (RMSE = 0.1474 rad, R<sup>2</sup> = 0.7865) under rapid regime shifts. Compared with CNN, LSTM, TCN, and Transformer baselines in both swell-dominated and typhoon-dominated regimes, the proposed method yields lower composite errors and more stable performance, with particularly robust directional reconstruction enabled by circular encoding. These results demonstrate strong accuracy, stability, and adaptability under sparse deployments, providing practical insights for observing-system design and multi-source data fusion in ocean engineering.</div></div>","PeriodicalId":19403,"journal":{"name":"Ocean Engineering","volume":"353 ","pages":"Article 124786"},"PeriodicalIF":5.5,"publicationDate":"2026-04-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147386354","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A rolling multi-objective ship speed optimization strategy for energy efficiency and operating cost control in dynamic waterway conditions 动态航道条件下船舶能效与运营成本的滚动多目标航速优化策略

IF 5.5 2区工程技术 Q1 ENGINEERING, CIVIL

Ocean Engineering

Pub Date : 2026-04-30 Epub Date: 2026-02-18 DOI: 10.1016/j.oceaneng.2026.124738

Yupeng Yuan , Xiaodong Guo , Zhe Xiong , Daogui Tang , Liang Tong , Chi Zhang , Zhengmao Li , Joshua Y. Kim

The push to decarbonization for ships is driving the development of supporting tools for operations. This study develops a rolling, voyage-scale multi-objective decision-support framework for speed management on inland waterways. The framework simultaneously minimizes the Energy Efficiency Operational Indicator (EEOI) and total operating cost by adjusting ship speed leg by leg under time-varying environmental conditions. Within each decision window, it solves a nonlinear bi-objective optimization problem using updated environmental and operating information to generate a practical compromise speed recommendation. A case study on a 7000 t bulk carrier across 18 legs shows that, compared with as-operated records, the proposed framework lowers total cost by 7.5% and shortens voyage time by 9.5%, while producing a smoother speed profile that is more compatible with ship propulsion and power-system control. These results indicate that rolling multi-objective speed optimization offers an effective engineering approach for voyage planning and operational speed management under time-varying environmental conditions, with clear potential for inland waterway decision-support applications.

船舶脱碳的推动正在推动运营配套工具的发展。本研究为内河航道航速管理开发了一个滚动、航次尺度的多目标决策支持框架。该框架通过在时变环境条件下逐段调整船舶速度，同时将能效运行指标（EEOI）和总运营成本降至最低。在每个决策窗口内，利用更新的环境和运行信息求解非线性双目标优化问题，生成实用的折衷速度建议。对一艘7000吨散货船18条腿的案例研究表明，与运行记录相比，所提出的框架降低了7.5%的总成本，缩短了9.5%的航行时间，同时产生了更平稳的航速剖面，与船舶推进和动力系统控制更兼容。这些结果表明，滚动多目标航速优化为时变环境下航次规划和航速管理提供了有效的工程方法，在内河航道决策支持应用中具有明显的潜力。

{"title":"A rolling multi-objective ship speed optimization strategy for energy efficiency and operating cost control in dynamic waterway conditions","authors":"Yupeng Yuan , Xiaodong Guo , Zhe Xiong , Daogui Tang , Liang Tong , Chi Zhang , Zhengmao Li , Joshua Y. Kim","doi":"10.1016/j.oceaneng.2026.124738","DOIUrl":"10.1016/j.oceaneng.2026.124738","url":null,"abstract":"<div><div>The push to decarbonization for ships is driving the development of supporting tools for operations. This study develops a rolling, voyage-scale multi-objective decision-support framework for speed management on inland waterways. The framework simultaneously minimizes the Energy Efficiency Operational Indicator (EEOI) and total operating cost by adjusting ship speed leg by leg under time-varying environmental conditions. Within each decision window, it solves a nonlinear bi-objective optimization problem using updated environmental and operating information to generate a practical compromise speed recommendation. A case study on a 7000 t bulk carrier across 18 legs shows that, compared with as-operated records, the proposed framework lowers total cost by 7.5% and shortens voyage time by 9.5%, while producing a smoother speed profile that is more compatible with ship propulsion and power-system control. These results indicate that rolling multi-objective speed optimization offers an effective engineering approach for voyage planning and operational speed management under time-varying environmental conditions, with clear potential for inland waterway decision-support applications.</div></div>","PeriodicalId":19403,"journal":{"name":"Ocean Engineering","volume":"353 ","pages":"Article 124738"},"PeriodicalIF":5.5,"publicationDate":"2026-04-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147386708","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A CNN-LSTM-Attention model for risk prediction of offshore wind turbines 海上风电机组风险预测的CNN-LSTM-Attention模型

IF 5.5 2区工程技术 Q1 ENGINEERING, CIVIL

Ocean Engineering

Pub Date : 2026-04-30 Epub Date: 2026-02-18 DOI: 10.1016/j.oceaneng.2026.124589

Jia-Xuan Li , Da-Gang Lu , Jia-Wei Ding , Yong Ding , Hao Liu , Libo Chen , Wenbing Wu

Offshore wind turbines are exposed to complex environmental loads such as wind, waves and ocean currents, which can lead to progressive structural degradation and increased failure risk over time. To achieve accurate risk prediction and intelligent structural health management, this paper proposes a hybrid deep learning model—CNN-LSTM-Attention—that integrates Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM) networks, and an Attention mechanism. The model leverages CNN to extract local spatial features, LSTM to capture temporal dynamics, and Attention to enhance the learning of critical patterns, thereby improving overall prediction accuracy. The proposed model is trained and validated on datasets representing four critical limit states of offshore wind turbine structures. Comparative experiments with traditional LSTM and CNN-LSTM models demonstrate that the CNN-LSTM-Attention model achieves superior performance across multiple evaluation metrics, including Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and the coefficient of determination (R²). Furthermore, the predicted probability distributions of structural failure reveal clear risk trends under different limit conditions, offering meaningful insights for risk-based maintenance planning and operational decision-making.

海上风力涡轮机暴露在复杂的环境载荷下，如风、波浪和洋流，这可能导致结构逐渐退化，并随着时间的推移增加故障风险。为了实现准确的风险预测和智能结构健康管理，本文提出了一种融合卷积神经网络（CNN）、长短期记忆（LSTM）网络和注意机制的混合深度学习模型CNN-LSTM-Attention。该模型利用CNN提取局部空间特征，利用LSTM捕捉时间动态，利用Attention增强对关键模式的学习，从而提高整体预测精度。所提出的模型在代表海上风力涡轮机结构四个临界极限状态的数据集上进行了训练和验证。与传统LSTM和CNN-LSTM模型的对比实验表明，CNN-LSTM-注意力模型在包括均方根误差（RMSE）、平均绝对误差（MAE）和决定系数（R2）在内的多个评价指标上都取得了优异的表现。此外，预测的结构失效概率分布揭示了不同极限条件下的风险趋势，为基于风险的维修计划和运营决策提供了有意义的见解。

{"title":"A CNN-LSTM-Attention model for risk prediction of offshore wind turbines","authors":"Jia-Xuan Li , Da-Gang Lu , Jia-Wei Ding , Yong Ding , Hao Liu , Libo Chen , Wenbing Wu","doi":"10.1016/j.oceaneng.2026.124589","DOIUrl":"10.1016/j.oceaneng.2026.124589","url":null,"abstract":"<div><div>Offshore wind turbines are exposed to complex environmental loads such as wind, waves and ocean currents, which can lead to progressive structural degradation and increased failure risk over time. To achieve accurate risk prediction and intelligent structural health management, this paper proposes a hybrid deep learning model—CNN-LSTM-Attention—that integrates Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM) networks, and an Attention mechanism. The model leverages CNN to extract local spatial features, LSTM to capture temporal dynamics, and Attention to enhance the learning of critical patterns, thereby improving overall prediction accuracy. The proposed model is trained and validated on datasets representing four critical limit states of offshore wind turbine structures. Comparative experiments with traditional LSTM and CNN-LSTM models demonstrate that the CNN-LSTM-Attention model achieves superior performance across multiple evaluation metrics, including Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and the coefficient of determination (R<sup>2</sup>). Furthermore, the predicted probability distributions of structural failure reveal clear risk trends under different limit conditions, offering meaningful insights for risk-based maintenance planning and operational decision-making.</div></div>","PeriodicalId":19403,"journal":{"name":"Ocean Engineering","volume":"353 ","pages":"Article 124589"},"PeriodicalIF":5.5,"publicationDate":"2026-04-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147386433","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A practical duct design to improve the performance of Darrieus hydrokinetic turbine kept in river flows 一种实用的水轮机保流风道设计

IF 5.5 2区工程技术 Q1 ENGINEERING, CIVIL

Ocean Engineering

Pub Date : 2026-04-30 Epub Date: 2026-02-23 DOI: 10.1016/j.oceaneng.2026.124801

S. Madhankumar , Atul Kumar Soti

This study investigates the performance enhancement of an H-type Darrieus vertical-axis hydrokinetic turbine augmented with a convergent-throat-divergent duct operating under realistic open-domain river flow conditions. Two-dimensional computational fluid dynamics simulations were performed using the SST

k - ω

turbulence model to accurately resolve the unsteady flow behaviour around the duct-turbine system. The effects of duct Contraction Ratio (CR) and throat length with different Tip Speed Ratios (TSRs) on the turbine performance were systematically analyzed. The results show that duct augmentation significantly improves energy extraction compared to the unducted turbine by accelerating the incoming flow and enhancing torque generation at the rotor plane. The ducted turbine achieves a maximum power coefficient (

C_{p}

) of 1.29 for a CR of 3.33 at TSR 2.5. Notably, the

C_{p}

exceeds the conventional Betz limit, which is permissible due to flow concentration effects induced by the duct in an open domain. At lower TSRs, intermediate CRs demonstrated better performance, highlighting the importance of matching duct geometry to flow conditions. Additional analyses of torque generation and velocity distribution confirm that the duct effectively accelerates flow through the rotor plane, enabling greater energy extraction. Variations in throat length show only a marginal impact on overall performance. Overall, these findings highlight that practical convergent-throat-divergent duct augmentation in an open-domain environment is a powerful strategy to enhance the turbine's performance in low-velocity river flow environments.

本文研究了h型Darrieus型垂直轴水动力水轮机在实际开域水流条件下的性能增强。采用SST k−ω湍流模型进行二维计算流体动力学模拟，以精确解析风管-涡轮系统周围的非定常流动行为。系统分析了不同叶尖速比下的风道收缩比和喉道长度对涡轮性能的影响。结果表明，与导流涡轮相比，增厚涡轮通过加速来流和增强转子平面的转矩产生，显著提高了能量提取。导管式涡轮在TSR为2.5时，CR为3.33，最大功率系数（Cp）为1.29。值得注意的是，Cp超过了传统的贝茨极限，这是允许的，这是由于管道在开放区域引起的流动浓度效应。在较低的tsr下，中间cr表现出更好的性能，突出了将管道几何形状与流动条件相匹配的重要性。对扭矩产生和速度分布的进一步分析证实，该管道有效地加速了通过转子平面的流动，从而实现了更大的能量提取。喉部长度的变化对整体性能的影响微乎其微。总的来说，这些发现突出表明，在开放域环境中，实际的收敛-喉部-发散风道增大是提高涡轮在低速河流环境中性能的有力策略。

{"title":"A practical duct design to improve the performance of Darrieus hydrokinetic turbine kept in river flows","authors":"S. Madhankumar , Atul Kumar Soti","doi":"10.1016/j.oceaneng.2026.124801","DOIUrl":"10.1016/j.oceaneng.2026.124801","url":null,"abstract":"<div><div>This study investigates the performance enhancement of an H-type Darrieus vertical-axis hydrokinetic turbine augmented with a convergent-throat-divergent duct operating under realistic open-domain river flow conditions. Two-dimensional computational fluid dynamics simulations were performed using the SST <span><math><mrow><mi>k</mi><mo>−</mo><mi>ω</mi></mrow></math></span> turbulence model to accurately resolve the unsteady flow behaviour around the duct-turbine system. The effects of duct Contraction Ratio (CR) and throat length with different Tip Speed Ratios (TSRs) on the turbine performance were systematically analyzed. The results show that duct augmentation significantly improves energy extraction compared to the unducted turbine by accelerating the incoming flow and enhancing torque generation at the rotor plane. The ducted turbine achieves a maximum power coefficient (<span><math><mrow><msub><mi>C</mi><mi>p</mi></msub></mrow></math></span>) of 1.29 for a CR of 3.33 at TSR 2.5. Notably, the <span><math><mrow><msub><mi>C</mi><mi>p</mi></msub></mrow></math></span> exceeds the conventional Betz limit, which is permissible due to flow concentration effects induced by the duct in an open domain. At lower TSRs, intermediate CRs demonstrated better performance, highlighting the importance of matching duct geometry to flow conditions. Additional analyses of torque generation and velocity distribution confirm that the duct effectively accelerates flow through the rotor plane, enabling greater energy extraction. Variations in throat length show only a marginal impact on overall performance. Overall, these findings highlight that practical convergent-throat-divergent duct augmentation in an open-domain environment is a powerful strategy to enhance the turbine's performance in low-velocity river flow environments.</div></div>","PeriodicalId":19403,"journal":{"name":"Ocean Engineering","volume":"353 ","pages":"Article 124801"},"PeriodicalIF":5.5,"publicationDate":"2026-04-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147386613","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Vulnerability analysis of offshore wind turbines considering the correlation of wind and waves 考虑风浪相关性的海上风力发电机组脆弱性分析

IF 5.5 2区工程技术 Q1 ENGINEERING, CIVIL

Ocean Engineering

Pub Date : 2026-04-30 Epub Date: 2026-02-18 DOI: 10.1016/j.oceaneng.2026.124636

Tao Wu , Bo Chen , Ying Sun

Vulnerability analysis, by considering load and structural uncertainties, provides an approach to reducing costs through optimized structural performance utilization. Addressing limitations in existing research regarding wind-wave correlation and performance level, this manuscript employs measured wind-wave data to estimate coherence functions using a 4th-order bandpass filter and 12th-order vector autoregressive. By analyzing wind-wave coherence under different sea states, a coherence function model is established and validated. OpenFAST is then utilized to generate wind-wave load time histories incorporating wind-wave correlations. An OWT numerical model is established in ANSYS to investigate dynamic responses, involving platform pitching, nacelle acceleration, and stress of tower and blade. A unified damage evaluation indicator and a three-stage performance level division standard for OWTs are proposed accordingly. The study subsequently conducts vulnerability analysis considering both wind-wave correlations and three-stage performance level. The study demonstrate that the proposed division standard effectively distinguishes between slight, moderate, and severe damage of OWT, with significant probability variations under identical load. This provides a reference for the design to fully utilize performance of OWTs. Notably, probabilities of achieving each performance level increase when wind-wave correlations are considered compared to uncorrelated cases, underscoring the necessity of incorporating wind-wave correlation in performance-based OWT design.

脆弱性分析通过考虑荷载和结构的不确定性，提供了通过优化结构性能利用来降低成本的方法。针对现有研究在风波相关性和性能水平方面的局限性，本文采用测量的风波数据，使用4阶带通滤波器和12阶矢量自回归估计相干函数。通过分析不同海况下的风浪相干性，建立并验证了相干函数模型。然后利用OpenFAST生成包含风波相关性的风波载荷时程。在ANSYS中建立了OWT数值模型，研究了包括平台俯仰、机舱加速度、塔架和叶片应力在内的OWT动力响应。据此提出了统一的损伤评价指标和三级性能等级划分标准。随后，考虑风浪相关性和三级性能水平，进行脆弱性分析。研究表明，所提出的划分标准能够有效区分轻型、中度和重度水轮机损伤，在相同荷载下，水轮机损伤概率变化显著。这为设计充分利用owt的性能提供了参考。值得注意的是，与不相关的情况相比，考虑风波相关性时，实现每个性能水平的概率都增加了，这强调了在基于性能的OWT设计中纳入风波相关性的必要性。

{"title":"Vulnerability analysis of offshore wind turbines considering the correlation of wind and waves","authors":"Tao Wu , Bo Chen , Ying Sun","doi":"10.1016/j.oceaneng.2026.124636","DOIUrl":"10.1016/j.oceaneng.2026.124636","url":null,"abstract":"<div><div>Vulnerability analysis, by considering load and structural uncertainties, provides an approach to reducing costs through optimized structural performance utilization. Addressing limitations in existing research regarding wind-wave correlation and performance level, this manuscript employs measured wind-wave data to estimate coherence functions using a 4th-order bandpass filter and 12th-order vector autoregressive. By analyzing wind-wave coherence under different sea states, a coherence function model is established and validated. OpenFAST is then utilized to generate wind-wave load time histories incorporating wind-wave correlations. An OWT numerical model is established in ANSYS to investigate dynamic responses, involving platform pitching, nacelle acceleration, and stress of tower and blade. A unified damage evaluation indicator and a three-stage performance level division standard for OWTs are proposed accordingly. The study subsequently conducts vulnerability analysis considering both wind-wave correlations and three-stage performance level. The study demonstrate that the proposed division standard effectively distinguishes between slight, moderate, and severe damage of OWT, with significant probability variations under identical load. This provides a reference for the design to fully utilize performance of OWTs. Notably, probabilities of achieving each performance level increase when wind-wave correlations are considered compared to uncorrelated cases, underscoring the necessity of incorporating wind-wave correlation in performance-based OWT design.</div></div>","PeriodicalId":19403,"journal":{"name":"Ocean Engineering","volume":"353 ","pages":"Article 124636"},"PeriodicalIF":5.5,"publicationDate":"2026-04-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147386347","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

ConvNeXt-PIV: A modern convolutional framework for particle image velocimetry ConvNeXt-PIV：一个用于粒子图像测速的现代卷积框架

IF 5.5 2区工程技术 Q1 ENGINEERING, CIVIL

Ocean Engineering

Pub Date : 2026-04-30 Epub Date: 2026-02-19 DOI: 10.1016/j.oceaneng.2026.124633

Zhifeng Xiao, Bo Hong, Liang Shan

In recent years, deep learning–based Particle Image Velocimetry (PIV) models have made remarkable progress in estimating velocity fields of complex flows. Conventional CNN-based PIV networks, relying on small convolutional kernels and pooling layers, are limited in capturing global structures and large-scale motions. Transformer-based models mitigate this limitation but incur substantial computational and memory costs. To address these issues, we propose ConvNeXt-PIV, an encoder built on the ConvNeXt module, where depthwise convolution serves as the main feature extractor and large kernels expand the receptive field. By redesigning the feature hierarchy and raising the output to one quarter of the input resolution, the network better captures particle spacing and velocity-gradient details. Experimental results on Problem Class I demonstrate that ConvNeXt-PIV educes the average endpoint error (AEE) by 50.0%, 68.6%, and 37.6% on the Backstep, JHTDB Channel, and DNS Turbulence datasets, respectively, compared with the baseline method. Meanwhile, the number of model parameters is reduced to 1.7M parameters while maintaining a comparable computational cost, highlighting a favorable balance between accuracy and efficiency. These results verify the potential of modern convolutional architectures for fluid-motion estimation and provide a new design paradigm for lightweight and high-precision PIV modeling.

近年来，基于深度学习的粒子图像测速（PIV）模型在复杂流体的速度场估计方面取得了显著进展。传统的基于cnn的PIV网络依赖于小卷积核和池化层，在捕获全局结构和大规模运动方面受到限制。基于变压器的模型减轻了这种限制，但会产生大量的计算和内存成本。为了解决这些问题，我们提出了ConvNeXt- piv，一种基于ConvNeXt模块的编码器，其中深度卷积作为主要的特征提取器，大核扩展接受域。通过重新设计特征层次并将输出分辨率提高到输入分辨率的四分之一，该网络可以更好地捕获粒子间距和速度梯度细节。在问题类I上的实验结果表明，与基线方法相比，ConvNeXt-PIV方法在Backstep、JHTDB Channel和DNS Turbulence数据集上的平均端点误差（AEE）分别降低了50.0%、68.6%和37.6%。同时，在保持相当的计算成本的同时，将模型参数数量减少到1.7M个参数，突出了精度和效率之间的良好平衡。这些结果验证了现代卷积架构在流体运动估计方面的潜力，并为轻量化和高精度PIV建模提供了新的设计范例。

{"title":"ConvNeXt-PIV: A modern convolutional framework for particle image velocimetry","authors":"Zhifeng Xiao, Bo Hong, Liang Shan","doi":"10.1016/j.oceaneng.2026.124633","DOIUrl":"10.1016/j.oceaneng.2026.124633","url":null,"abstract":"<div><div>In recent years, deep learning–based Particle Image Velocimetry (PIV) models have made remarkable progress in estimating velocity fields of complex flows. Conventional CNN-based PIV networks, relying on small convolutional kernels and pooling layers, are limited in capturing global structures and large-scale motions. Transformer-based models mitigate this limitation but incur substantial computational and memory costs. To address these issues, we propose ConvNeXt-PIV, an encoder built on the ConvNeXt module, where depthwise convolution serves as the main feature extractor and large kernels expand the receptive field. By redesigning the feature hierarchy and raising the output to one quarter of the input resolution, the network better captures particle spacing and velocity-gradient details. Experimental results on Problem Class I demonstrate that ConvNeXt-PIV educes the average endpoint error (AEE) by 50.0%, 68.6%, and 37.6% on the Backstep, JHTDB Channel, and DNS Turbulence datasets, respectively, compared with the baseline method. Meanwhile, the number of model parameters is reduced to 1.7M parameters while maintaining a comparable computational cost, highlighting a favorable balance between accuracy and efficiency. These results verify the potential of modern convolutional architectures for fluid-motion estimation and provide a new design paradigm for lightweight and high-precision PIV modeling.</div></div>","PeriodicalId":19403,"journal":{"name":"Ocean Engineering","volume":"353 ","pages":"Article 124633"},"PeriodicalIF":5.5,"publicationDate":"2026-04-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147387252","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Novel algorithm for the calibration of DVL in underwater integrated navigation system 水下组合导航系统DVL标定的新算法

IF 5.5 2区工程技术 Q1 ENGINEERING, CIVIL

Ocean Engineering

Pub Date : 2026-04-30 Epub Date: 2026-02-21 DOI: 10.1016/j.oceaneng.2026.124676

Haixu Zhang , Chong Li , Tao Zhang , Guangcai Wang , Di Wang

Calibration plays a critical role in enhancing the performance of underwater integrated navigation systems. Widely used calibration algorithms based on filtering theory often neglect higher-order disturbance terms in the expansion of the target equation and commonly assume misalignment angles as small quantities for approximate calculations. In engineering applications, when misalignment angles cannot be considered negligible, the resulting computational errors become significant.

To address this problem, we propose a novel Local Area Segmentation (LAS) optimization algorithm for Doppler Velocity Log (DVL) error calibration that incorporates all error terms. Building upon the commonly used single-valued scale factor calibration model, we develop extended calibration models to include three-axis scale factors and four-beam scale factors. All three models are utilized to validate the superiority of the proposed algorithm. Experimental results demonstrate that the proposed algorithm achieves higher estimation accuracy, significantly improving DVL calibration precision.

标定对提高水下组合导航系统的性能起着至关重要的作用。常用的基于滤波理论的标定算法在目标方程展开时往往忽略了高阶扰动项，通常将不对准角作为小量进行近似计算。在工程应用中，当不对准角不能被认为是可以忽略的时候，由此产生的计算误差就会变得很大。为了解决这个问题，我们提出了一种新的局部区域分割（LAS）优化算法，用于多普勒速度日志（DVL）误差校准，该算法包含所有误差项。在常用的单值比例因子校准模型的基础上，我们开发了扩展的包括三轴比例因子和四梁比例因子的校准模型。通过三个模型验证了所提算法的优越性。实验结果表明，该算法具有较高的估计精度，显著提高了DVL标定精度。

{"title":"Novel algorithm for the calibration of DVL in underwater integrated navigation system","authors":"Haixu Zhang , Chong Li , Tao Zhang , Guangcai Wang , Di Wang","doi":"10.1016/j.oceaneng.2026.124676","DOIUrl":"10.1016/j.oceaneng.2026.124676","url":null,"abstract":"<div><div>Calibration plays a critical role in enhancing the performance of underwater integrated navigation systems. Widely used calibration algorithms based on filtering theory often neglect higher-order disturbance terms in the expansion of the target equation and commonly assume misalignment angles as small quantities for approximate calculations. In engineering applications, when misalignment angles cannot be considered negligible, the resulting computational errors become significant.</div><div>To address this problem, we propose a novel Local Area Segmentation (LAS) optimization algorithm for Doppler Velocity Log (DVL) error calibration that incorporates all error terms. Building upon the commonly used single-valued scale factor calibration model, we develop extended calibration models to include three-axis scale factors and four-beam scale factors. All three models are utilized to validate the superiority of the proposed algorithm. Experimental results demonstrate that the proposed algorithm achieves higher estimation accuracy, significantly improving DVL calibration precision.</div></div>","PeriodicalId":19403,"journal":{"name":"Ocean Engineering","volume":"353 ","pages":"Article 124676"},"PeriodicalIF":5.5,"publicationDate":"2026-04-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147387096","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0