船舶柴油机是船舶动力的来源，同时也是保障船舶安全航行的重要保证。柴油机涡轮增压器故障存在难以直接识别、故障数量多、多种故障共存等特点，传统的故障诊断难以解决涡轮增压器的故障诊断问题。本文基于BP神经网络建立柴油机涡轮增压器故障诊断模型，分析BP神经网络的基本结构，阐述故障诊断模型的基本工作过程，在建立的故障诊断模型基础上对不同故障进行识别，并研究不同隐含层数量时对故障识别正确率的影响。研究发现，隐含层数量等于14时故障识别正确率最高。本文建立的船舶柴油机涡轮增压器故障诊断系统依托于实际故障数据样本，测试结果表明该故障诊断模型能够有效应用于涡轮增压器的故障诊断工作。
Marine diesel engine is the source of marine power, but also an important guarantee to ensure the safe navigation of the ship. The fault diagnosis of diesel turbocharger is difficult to identify directly, has a large number of faults, and many faults coexist, so the traditional fault diagnosis is difficult to solve the problem of turbocharger fault diagnosis. This paper establishes a fault diagnosis model of diesel turbocharger based on BP neural network, analyzes the basic structure of BP neural network, describes the basic working process of the fault diagnosis model, identifies different faults based on the established fault diagnosis model, and studies the influence of different hidden layers on the fault identification accuracy rate. When the number of hidden layers is equal to 14, the correct rate of fault identification is the highest. The fault diagnosis system of Marine diesel turbocharger established in this paper is based on actual fault data samples, and the test results show that the fault diagnosis model can be effectively applied to the fault diagnosis work of turbocharger.
2023,45(20): 155-158 收稿日期：2023-4-10
DOI：10.3404/j.issn.1672-7649.2023.20.029
分类号：U664.121
基金项目：浙江省中华职业教育科研项目课题(ZJCV2023E15)
作者简介：袁对(1979-),男,硕士,讲师,研究方向为船舶轮机工程技术
参考文献：
[1] 石大亮, 张毅然, 湛日景, 等. 基于关联规则分类的船用柴油机故障诊断[J]. 内燃机学报, 2023, 41(4): 369-375.
SHI Da-liang, ZHANG Yi-ran, ZHAN Ri-jing, et al. Fault diagnosis of Marine Diesel Engines based on association rule classification[J]. Transactions of CSICE, 2023, 41(4): 369-375.
[2] 曹乐乐, 张鹏, 高泽宇, 等. 基于动态特征融合的船舶柴油机进排气系统故障诊断[J]. 中国舰船研究, 2022, 17(6): 103-110.
CAO Le-le, ZHANG Peng, GAO Ze-yu, et al. Fault diagnosis of Marine diesel Engine intake and Exhaust system based on Dynamic Feature fusion[J]. Chinese Ship Research, 2022, 17(6): 103-110.
[3] 邱其清, 廖志强. 基于高斯混合和概率神经网络的舰船柴油机故障诊断方法[J]. 船舶工程, 2022, 44(9): 101-106+113.
QIU Qi-Qing, LIAO Zhi-qiang. Fault diagnosis method of Marine diesel engine based on Gaussian mixture and probabilistic neural network[J]. Marine Engineering, 2022, 44(9): 101-106+113.
[4] 张靖. 基于数据驱动的船舶柴油机智能故障诊断研究[J]. 内燃机与配件, 2022(19): 79-81.
ZHANG Jing. Research on intelligent fault diagnosis of marine diesel engine based on data drive[J]. Internal Combustion Engine and Accessories, 2022(19): 79-81.
[5] 胡继敏,罗梅杰. 基于自监督学习框架的发电柴油机故障诊断[J]. 船电技术, 2022, 42(9): 19-24.
HU Ji-min, LUO Mei-jie. Fault diagnosis of power generation diesel Engine based on Self-supervised Learning framework[J]. Marine Electric Technology, 2022, 42(9): 19-24.
[6] 石新发, 邢广笑, 贺石中, 等. 基于熵理论和BP神经网络的船舶柴油机磨损故障识别[J]. 润滑与密封, 2022, 47(7): 54-58.
SHI Xin-fa, XING Guang-xiao, HE Shi-zhong, et al. Wear fault Identification of marine diesel engine based on entropy theory and BP neural network[J]. Lubrication & Seals, 2022, 47(7): 54-58.
