为保障舰船安全航行，研究基于粒子群优化算法的舰船故障点智能定位方法。采集舰船机械设备运行过程中的振动信号，根据振动信号判断舰船机械设备运行工况是否正常。针对工况异常的振动信号，采用基于小波包理论，根据异常振动信号的多分辨率转换对其实施正交分解，并在此基础上提取特征向量；依照异常振动信号间自熵和互熵的相关性划分特征向量类别。将各类舰船设备振动信号特征向量作为BP神经网络的输入，利用粒子群优化算法全局搜索BP神经网络最优权值与阈值，并通过确定粒子反向适应度提升参数搜索效率，由此优化BP神经网络模型结构，完成舰船故障识别以及故障点智能定位目的。实验结果显示所研究方法在单一故障或多种故障并存条件下就能够准确定位机械设备故障点，同时准确识别机械设备故障，保障舰船安全航行。
In order to ensure the safe navigation of ships, the intelligent location method of ship fault points based on particle swarm optimization algorithm is studied. Collect the vibration signal during the operation of ship mechanical equipment, and judge whether the operation condition of ship mechanical equipment is normal according to the vibration signal. According to the multi-resolution transformation of abnormal vibration signal based on wavelet packet theory, the orthogonal decomposition is implemented, and the feature vector is extracted on this basis; The eigenvectors are classified according to the correlation between self entropy and mutual entropy between abnormal vibration signals. Taking the vibration signal characteristics of ship equipment as the input of BP neural network model, the particle swarm optimization algorithm is used to globally search the optimal weight and threshold of BP neural network, and the parameter search efficiency is improved by determining the particle reverse fitness, so as to optimize the structure of BP neural network model and complete the purpose of ship fault identification and intelligent fault location. The experimental results show that the research method can accurately locate the fault point of mechanical equipment and accurately identify the fault of mechanical equipment under the condition of single fault or multiple faults.
2022,44(10): 159-162 收稿日期：2021-11-24
DOI：10.3404/j.issn.1672-7649.2022.10.034
分类号：TP206
作者简介：甘杜芬(1984-)，女，硕士，高级工程师，研究方向为WEB前端开发及算法
参考文献：
[1] 王瑞涵, 陈辉, 管聪. 基于机器学习的船舶机舱设备状态监测方法[J]. 中国舰船研究, 2021, 16(1): 158–167
[2] 吴旭升, 杨刚, 孙盼, 等. 舰船电力电子设备测试性设计现状及技术分析[J]. 海军工程大学学报, 2021, 33(2): 14–18
[3] 孙宇嫣, 蔡泽祥, 郭采珊, 等. 基于深度学习的智能变电站通信网络故障诊断与定位方法[J]. 电网技术, 2019, 43(12): 4306–4314
[4] 周光宇, 马松龄. 基于机器学习与DGA的变压器故障诊断及定位研究[J]. 高压电器, 2020, 56(6): 262–268
[5] 赵传宗, 王中杰, 穆景龙, 等. 基于小波变换和遗传算法的小电流接地故障定位技术研究[J]. 工业安全与环保, 2019, 45(1): 62–65
[6] 郦阳, 王宝华. 继电保护系统故障的智能定位方法研究[J]. 电力系统保护与控制, 2022, 50(2): 69–76
[7] 刘宇涛, 孙虎儿. 基于粒子群优化的CYCBD在滚动轴承故障特征提取的应用研究[J]. 机械传动, 2021, 45(2): 171–176
[8] 张兰勇, 孟坤, 刘胜, 等. 基于改进双粒子群算法的舰船电力系统网络故障重构[J]. 电力系统保护与控制, 2201, 9,47(9): 90–96
[9] 印波, 王锡淮, 肖健梅. 基于改进粒子群优化算法的船舶能量管理方案[J]. 中国舰船研究, 2020, 15(6): 37–45
