基于LRDP的转子故障数据集降维方法

兰州理工大学学报 ›› 2026, Vol. 52 ›› Issue (2): 55-62.

基于LRDP的转子故障数据集降维方法

梁启博¹, 赵荣珍^*1,2, 邓林峰¹

1.兰州理工大学机电工程学院, 甘肃兰州 730050;
2.广州理工学院智能制造与电气工程学院, 广东广州 510540

收稿日期:2023-04-19 出版日期:2026-04-28 发布日期:2026-04-28
通讯作者: 赵荣珍(1960-),女,山东枣庄人,博士,教授,博导. Email:zhaorongzhen@lut.edu.cn
基金资助:
国家自然科学基金(62241308)

Dimension reduction method of rotor fault data set based on LRDP

LIANG Qi-bo¹, ZHAO Rong-zhen^1,2, DENG Lin-feng¹

1. School of Mechanical and Electrical Engineering, Lanzhou University of Technology, Lanzhou 730050, China;
2. College of Intelligent Manufacturing and Electrical Engineering, Guangzhou Institute of Science and Technology, Guangzhou 510540, China

Received:2023-04-19 Online:2026-04-28 Published:2026-04-28

摘要/Abstract

摘要： 针对高维故障数据集导致故障辨识效果不佳的问题,提出了基于局部重构与边界判别集成的多图协同转子故障数据集降维算法.首先,从多域提取转子振动信号的故障特征,并构建高维故障数据集;其次,利用各近邻样本线性重构原始样本,从而保持故障数据的内在流形结构;最后,基于重构的样本点构建故障特征数据集的本征图、惩罚图和判别图,并利用多图嵌入的方式进行降维.同时,采用2个高维转子故障数据集对算法的性能进行验证,结果表明,相比其它流形算法该算法具有更高的识别准确率和稳定性,总体识别精度可分别达到99.7%和98.7%.

关键词: 故障诊断, 图嵌入, 数据可视化, 数据降维, 局部重构

Abstract: To improve the poor fault identification precision caused by high-dimensional fault datasets, a multigraph cooperative rotor fault dataset dimensionality reduction strategy integrating local reconstruction and border discrimination is proposed. In order to maintain the intrinsic manifold structure of the fault data, the fault features of rotor vibration signals are extracted first from multiple domains and constructs a high-dimensional fault dataset. Next, each sample is linearly reconstructed using its neighboring samples. Finally, an eigenmap and penalty map,as well as discriminant map of the fault feature dataset are constructed, and dimensionality reduction is performed through a multi-graph embedding model. The algorithm's performance was evaluated using two distinct high-dimensional rotor failure datasets. The results demonstrated that the proposed algorithm surpassed other well-known algorithms in terms of recognition accuracy and stability. It has overall recognition accuracy of 99.7% and 98.7%, respectively.

Key words: fault diagnosis, graph embedding, visualization of data, dimension reduction, local remodeling

中图分类号:

TH165
TP18

梁启博, 赵荣珍, 邓林峰. 基于LRDP的转子故障数据集降维方法[J]. 兰州理工大学学报, 2026, 52(2): 55-62.

LIANG Qi-bo, ZHAO Rong-zhen, DENG Lin-feng. Dimension reduction method of rotor fault data set based on LRDP[J]. Journal of Lanzhou University of Technology, 2026, 52(2): 55-62.

导出引用管理器 EndNote|Reference Manager|ProCite|BibTeX|RefWorks

链接本文: https://journal.lut.edu.cn/CN/

https://journal.lut.edu.cn/CN/Y2026/V52/I2/55

参考文献

[1] JIANG F,ZHU Z C,LI W,et al.Fault identification of rotor bearing system based on ensemble empirical mode decomposition and self-zero space projection analysis [J].Journal of Sound and Vibration,2014,333(14):3321-3331.
[2] SU Z Q,TANG B P,MA J H,et al.Fault diagnosis method based onincremental enhanced supervised locally linear embedding and adaptive nearest neighbor classifier [J].Measurement,2014,48(1):136-148.
[3] 雷亚国,贾峰,孔德同,等.大数据下机械智能故障诊断的机遇与挑战 [J].机械工程学报,2018,54(5):94-104.
[4] HE X F,NIYOGI P.Locality preserving projections [J].Advances in Neural Information Processing Systems,2003,16:153-160.
[5] YAN S C,XU D,ZHANG B Y,et al.Graph embedding and extensions:a general framework for dimensionality reduction [J].IEEE Transcations on Pattern Analysis and Machine Intelligence,2007,29(1):40-51.
[6] ZHAO H T,SUN S Y,JING Z L,et al.Local structure based supervised feature extraction [J].Pattern Recognition,2006,39(8):1546-1550.
[7] 石明宽,赵荣珍.基于局部平衡判别投影的旋转机械故障诊断方法 [J].中国机械工程,2021,32(14):1653-1658.
[8] 刘嘉敏,杨松,黄鸿.基于局部重构Fisher分析的高光谱遥感影像分类 [J].中国激光,2020,47(7):390-401.
[9] 常书源,赵荣珍,陈博,等.基于边界判别多流形分析的故障数据集降维方法 [J].振动与冲击,2021,40(23):120-126.
[10] HUANG C B,ABEO T A,LUO X Z,et al.Semi-supervised manifold alignment with multi-graph embedding [J].Multimedia Tools and Applications,2020,79(27/28):20241-20262.
[11] 李勇智,杨静宇,郑宇杰,等.基于最大间距准则(MMC)新的有效特征提取方法 [J].系统仿真学报,2007(5):1061–1066.
[12] 周宏飞,赵荣珍.基于GSRDPGE算法的转子故障数据集降维方法研究 [J].兰州理工大学学报,2023,49(4):35-41.
[13] ZHAO X L,JIA M P.Fault diagnosis of rolling bearing on feature reduction with global-local margin fisher analysis [J].Neurocomputing,2018,315:447-464.
[14] 户文刚,赵荣珍.基于转子故障数据集的KSELF降维方法 [J].振动、测试与诊断,2021,41(1):13-18.

[1]	郭俊锋, 陈大朋, 王淼生, 王二化. 基于方差贡献率的齿轮箱裂纹故障诊断[J]. 兰州理工大学学报, 2025, 51(6): 42-48.
[2]	王江, 剡昌锋, 卢家伟, 王瑞民, 张永明. 融合多头自注意力机制的故障命名实体识别[J]. 兰州理工大学学报, 2025, 51(4): 33-42.
[3]	马宁, 赵荣珍, 郑玉巧. 基于CEEMDAN与改进一维多尺度卷积神经网络结合的滚动轴承故障诊断[J]. 兰州理工大学学报, 2025, 51(1): 45-54.
[4]	郑玉巧, 李浩, 魏泰. 基于OVMD-RF方法的风力发电机滚动轴承故障诊断[J]. 兰州理工大学学报, 2024, 50(4): 36-42.
[5]	张斌, 程珩, 孟倩. 基于ICEEMDAN和Hilbert包络谱的滚动轴承故障诊断研究[J]. 兰州理工大学学报, 2024, 50(3): 51-57.
[6]	杜先君, 邱小彧. 基于改进LSTM神经网络的化工过程故障诊断[J]. 兰州理工大学学报, 2023, 49(6): 72-79.
[7]	郭俊锋, 孙磊, 王淼生, 续德锋. 基于1D-LeNet-5模型的滚动轴承故障诊断方法[J]. 兰州理工大学学报, 2023, 49(5): 34-41.
[8]	赵小强, 罗维兰, 梁浩鹏. 基于SIR多级残差连接密集网络的轴承故障诊断[J]. 兰州理工大学学报, 2022, 48(6): 46-54.
[9]	朱奇先, 梁浩鹏, 赵小强, 宋昭様. 基于残差神经网络的滚动轴承故障诊断方法[J]. 兰州理工大学学报, 2022, 48(3): 86-93.
[10]	赵荣珍, 薛勇, 吴耀春. 改进的D-t-SNE滚动轴承故障数据集降维方法[J]. 兰州理工大学学报, 2022, 48(3): 42-49.
[11]	赵荣珍, 常书源. 基于M-WDLS和PCA的转子故障诊断方法[J]. 兰州理工大学学报, 2021, 47(5): 38-44.
[12]	赵荣珍, 马森财, 吴耀春. 云模型和集成极限学习机相结合的滚动轴承故障诊断方法[J]. 兰州理工大学学报, 2021, 47(4): 33-39.
[13]	赵荣珍, 陈昱吉. KPCA与LTSA融合的转子故障数据集降维算法[J]. 兰州理工大学学报, 2021, 47(1): 36-40.
[14]	陈鹏, 赵小强, 朱奇先. 基于VMD-MPE-KPCA特征提取与MRVM相混合的滚动轴承故障诊断方法[J]. 兰州理工大学学报, 2020, 46(5): 92-99.
[15]	剡昌锋, 栗宇, 王慧滨, 张强, 艾科勇, 吴黎晓. 基于本体的汽轮发电机组故障诊断知识建模[J]. 兰州理工大学学报, 2020, 46(5): 41-48.