电子测量与仪器学报

刘林凡,周丽芸.参数优化CNN-BiLSTM-Attention的船舶火灾风险评估方法[J].电子测量与仪器学报,2026,40(3):240-249

参数优化CNN-BiLSTM-Attention的船舶火灾风险评估方法

Parameter-optimized CNN-BiLSTM-Attention approach forship fire identification

DOI：

英文关键词:fire identification ship attention mechanism bidirectional long short-term memory grey wolf optimization

基金项目:江西省教育厅科学技术研究项目(GJJ2400501)、江西省自然基金项目（20252BAB25057）资助

作者	单位
刘林凡	1.华东交通大学电气与自动化工程学院南昌330000；2.华东交通大学控制科学与工程流动站南昌330000
周丽芸	江西工业贸易职业技术学院信息工程系南昌330000

Author	Institution
Liu Linfan	1.School of Electrical and Automation Engineering, East China Jiaotong University, Nanchang 330000, China; 2.Postdoctoral Research Station of Control Science and Engineering, East China Jiaotong University, Nanchang 330000, China
Zhou Liyun	Department of Information Engineering, Jiangxi Industry & Trade Vocational and Technical College, Nanchang 330000, China

摘要点击次数: 319

全文下载次数: 201

中文摘要:

针对现有船舶火灾辨识准确率不高问题，提出了一种融合注意力机制、卷积神经网络（convolutional neural network，CNN）和双向长短期记忆网络（bidirectional long short term memory，BiLSTM）的船舶火灾辨识方法。使用火灾动力学模拟工具（fire dynamics simulator，FDS）建立了3层甲板渡轮火灾仿真模型，利用传感器采集了船舶火灾过程的温度、一氧化碳浓度和能见度数据。随后采用CNN提取火灾数据的纵向特征，并通过降维压缩数据长度，减少模型的训练参数；搭建基于BiLSTM的级联深度学习神经网络，提取火灾数据横向特征，并在输出层融合注意力机制。此外，为了提升灰狼优化算法的收敛速度，将混沌博弈算法引入该算法，提出了改进的灰狼优化算法。最后，采用改进的灰狼优化算法优化CNN-BiLSTM-Attention模型，并利用该模型进行了两个场景的船舶火灾辨识实验。结果表明，相比于其他火灾分类方法，在船舶火灾数据样本不均衡情况下，所提方法火灾辨识的准确率达到100%，能够满足工程实际需求。

英文摘要:

This study proposed a ship fire identification approach that integrates an attention mechanism, a convolutional neural network (CNN) and a bidirectional long short term memory network (BiLSTM) to address the low accuracy of existing methods. A three-deck ferry fire simulation model was constructed using the fire dynamics simulator (FDS), and sensors were used to collect temperature, carbon monoxide, and visibility data from the simulated ship fire process. A CNN was employed to extract longitudinal features from fire data, while dimensionality reduction was used to compress data length and to reduce the number of model training parameters. A cascaded deep learning neural network based on BiLSTM was established to extract transverse features from fire data, where an attention mechanism was incorporated at the output layer. Furthermore, to accelerate convergence, an improved grey wolf optimization algorithm was developed by integrating the chaotic game algorithm. The improved algorithm was applied to optimize the CNN-BiLSTM-Attention model, which was subsequently utilized to perform ship fire identification experiments under two scenarios. The experimental results indicated that, despite the imbalance in ship fire data samples, the proposed approach outperformed other fire classification methods, achieving 100% fire identification accuracy and satisfying practical engineering requirements.

查看全文查看/发表评论下载PDF阅读器