为了精准识别电磁阀故障，确保船用柴油机安全、平稳运行，提出基于小波包分解的船用柴油机燃油电磁阀故障诊断方法。采用小波包分解法对船用柴油机燃油电磁阀电流信号作分解，获取其多频带特征。通过核主成分分析法对其作降维处理，完成敏感特征选择。将其作为最小支持向量机的输入，自适应蚁群优化算法通过自适应调整挥发因子、状态转移规则确定最优模型参数，实现燃油电磁阀故障的准确诊断。结果表明：故障、正常工况下的燃油电磁阀电流特性曲线存在较大差异；该方法可提取电流信号的8个频带特征、不同频带特征间差异度大；特征选择有利于提高燃油电磁阀故障辨识度。本文方法可实现燃油电磁阀故障诊断，诊断效果突出。
In order to accurately identify solenoid valve faults and ensure the safe and smooth operation of marine diesel engines, a fault diagnosis method for marine diesel fuel solenoid valves based on Wavelet packet decomposition is proposed. The Wavelet packet decomposition method is used to decompose the current signal of the marine diesel fuel solenoid valve to obtain its multi band characteristics, and the Kernel principal component analysis method is used to reduce its dimensions to complete the sensitive Feature selection, which is used as the input of the minimum support vector machine. The adaptive ant colony optimization algorithm determines the optimal model parameters by adaptively adjusting the volatilization factor and state transition rules to achieve accurate fault diagnosis of the fuel solenoid valve. The experimental results indicate that there are significant differences in the current characteristic curves of the fuel solenoid valve under fault and normal operating conditions. This method can extract 8 frequency band features of current signals, with significant differences between different frequency band features. Feature selection is conducive to improving the fault identification of fuel solenoid valve. It can achieve fault diagnosis of fuel solenoid valves, with outstanding diagnostic effects.
2023,45(18): 105-108 收稿日期：2023-06-17
DOI：10.3404/j.issn.1672-7649.2023.18.017
分类号：TH212
作者简介：葛君超(1988-),女,硕士,讲师,研究方向为塑性变形和断裂及产品结构优化
参考文献：
[1] 黄烨鑫, 万振刚, 程琛. 基于改进鲸鱼算法寻优SVM的船用柴油机燃油系统故障诊断方法研究[J]. 计算技术与自动化, 2021, 40(2): 53-56.
[2] 谢海龙, 陈国顺, 谢海勇, 等. 电磁阀常见故障的分析及处理[J]. 阀门, 2020, 230(4): 47-50.
[3] 张明兴. 发动机OCV电磁阀卡滞问题解析[J]. 汽车制造业, 2022, 677(1): 39-41.
[4] 马栋, 刘志浩, 高钦和, 等. 基于多特征融合的电磁换向阀故障模式识别[J]. 北京航空航天大学学报, 2023, 49(4): 913-921.
[5] 郝德琛, 李华玲, 黄晋英. 小波包分解和改进ResNet行星齿轮箱故障诊断方法[J]. 传感器与微系统, 2022, 41(8): 116-119+123.
[6] 王钦惠, 胡向宇, 崔梧玉, 等. 基于小波分析的电磁阀在轨实时诊断[J]. 真空与低温, 2021, 27(3): 292-295.
[7] 张文啸, 孟国香, 叶骞. 基于Triplet loss的电磁阀故障识别方法[J]. 液压与气动, 2022, 46(9): 116-125.
[8] 朱兴统. 基于小波包分解和K最近邻算法的轴承故障诊断方法[J]. 装备制造技术, 2020, 302(2): 24-27+45.
