船舶结构庞大且复杂，简单的分类模型难以满足船体损伤检测的全面需求。本文提出一种基于卷积神经网络和长短期记忆网络的船体板架损伤位置检测方法。通过Abaqus脚本对带有裂纹的船体板架进行参数化建模和响应计算，建立不同损伤位置板架的加速度响应数据集。然后，构建CNN-LSTM神经网络代理模型，直接以多通道加速度响应作为输入，避免了传统机器学习方法需要人工设计损伤敏感特征的问题，以裂纹中心坐标作为输出，避免了分类方法对损伤位置覆盖不充分的问题。结果显示，预测裂纹位置与实际位置的误差不超过板架尺寸的5%，噪声水平越低、裂纹越严重，此方法的预测精度越高，为船舶结构健康监测提供了较好的思路和方案。
The structure of ships is large and complex, and simple classification models are difficult to meet the comprehensive needs of hull damage detection. In this paper, a method for detecting the damage position of ship hull plate frames based on CNN and LSTM is proposed. The Abaqus script is used to parameterize the modeling and response calculation of ship hull frames with cracks, and the acceleration response data set of plate frames at different damage positions is established. Then, the CNN-LSTM neural network surrogate model is constructed, which directly takes the multi-channel acceleration responses as the input, which avoids the problem that the traditional machine learning method needs to manually design damage-sensitive features, and uses the crack center coordinates as the output, which avoids the problem that the classification method does not cover the damage location sufficiently. The results show that, the error between the predicted crack location and the actual position is not more than 5% of the plate frame size. The lower the noise level and the more severe the cracks, the higher the prediction accuracy of this method, which provides a better idea and scheme for ship structural health monitoring.
2025,47(19): 87-93 收稿日期：2024-10-18
DOI：10.3404/j.issn.1672-7649.2025.19.014
分类号：U661.44
作者简介：王文婷(2000-),女,硕士研究生,研究方向为结构损伤识别
参考文献：
[1] 张志民. 散货船船体结构损坏分析及强度评估[D]. 上海: 上海海事大学, 2006.
[2] AVCI O, ABDELJABER O, KIRANYAZ S, et al. A review of vibration-based damage detection in civil structures: from traditional methods to machine learning and deep learning applications[J]. Mechanical Systems and Signal Processing, 2021, 147: 107077.
[3] 腾帅. 基于深度学习的结构损伤识别方法研究[D]. 广州: 广东工业大学, 2020.
[4] ABDELJABER O, AVCI O, KIRANYAZ S, et al. Real-time vibration-based structural damage detection using one-dimensional convolutional neural networks[J]. Journal of Sound and Vibration, 2017, 388: 154-170.
[5] 张松林, 马栋梁, 王德禹. 基于长短期记忆神经网络的板裂纹损伤检测方法[J]. 上海交通大学学报, 2021, 55(5): 527-535.
ZHANG S L, MA D L, WANG D Y. Method for plate crack damage detection based on long short-term memory neural network[J]. Journal of Shanghai Jiao Tong University, 2021, 55(5): 527-535.
[6] MA D L, WANG D Y. A deep learning-based method for hull stiffened plate crack detection[J]. Proceedings of the Institution of Mechanical Engineers, Part M: Journal of Engineering for the Maritime Environment, 2021, 235(2): 570-585.
[7] 张健飞, 黄朝东, 王子凡. 基于多头自注意力机制和卷积神经网络的结构损伤识别研究[J]. 振动与冲击, 2022, 41(24): 60-71.
ZHANG J F, HUANG C D, WANG Z F. Research on structural damage identification based on multi-head self-attention and convolutional neural networks[J]. Journal of Vibration and Shock, 2022, 41(24): 60-71.
[8] 胡钧剑. 基于深度学习理论的桥梁结构损伤识别研究[D]. 南昌: 华东交通大学, 2021.
[9] 包均. 含裂纹和腐蚀船舶结构极限强度计算方法及系统可靠性研究[D]. 镇江: 江苏科技大学, 2018.
[10] ERNY C, THEVENET D, COGNARD J Y, et al. Fatigue life prediction of welded ship details[J]. Marine Structures, 2012, 25(1): 13-32.
[11] LIN Y, NIE Z, MA H. Structural damage detection with automatic feature-extraction through deep learning[J]. Computer-Aided Civil and Infrastructure Engineering, 2017, 32(12): 1025-1046.
