摘要
对船舶轴承信号进行实时监测可以有效提升船舶运行的安全和可靠性,本文提出一种船舶轴承信号监测系统,对监测系统的硬件结构框架进行设计,阐述各模块的功能实现原理,给出轴承传感器的安装结构设计方案,对某型号的船舶轴承不同状态下轴承振动信号进行测试,分析异常信号的类型和基本特征,在此基础上提出基于深度学习的船舶轴承信号特征提取和分类的流程,论述自注意力机制的实现过程。本文提出的船舶轴承信号监测系统不仅可以保证高实时性,同时可以完成轴承振动信号的准确判别。
Real-time monitoring of ship bearing signals can effectively improve the safety and reliability of ship operation.This paper proposes a ship bearing signal monitoring system,designs the hardware structure framework of the monitoring system,expounds the functional realization principle of each module,and gives the installation structure design scheme of the bearing sensor.The vibration signals of a certain type of ship bearing under different states are tested,and the types and basic characteristics of abnormal signals are analyzed.On this basis,the process of feature extraction and classification of ship bearing signals based on deep learning is proposed,and the realization process of self-attention mechanism is discussed.The ship bearing signal monitoring system proposed in this paper can not only ensure high real-time performance,but also accurately distinguish the bearing vibration signal.
作者
罗磊
李嵩
何韦玲
邹国仁
LUO Lei;LI Song;HE Weiling;ZOU Guoren(Wuhan University,Wuhan 430072,China;Jiangxi University of Software Professional Technology,Nanchang 330041,China)
出处
《舰船科学技术》
北大核心
2024年第22期155-158,共4页
Ship Science and Technology
基金
教育部产学合作协同育人项目(230705115235222)
江西省教育厅科学技术研究项目(GJJ2204614)。
关键词
轴承信号
深度学习
传感器
监测系统
振动信号特征
bearing signal
deep learning
sensor
monitoring system
vibration signal feature
