融合特征下的双流CNN的制动蠕动颤振评价

PDF(3301 KB)

振动与冲击 ›› 2025, Vol. 44 ›› Issue (1) : 134-142.

故障诊断分析

融合特征下的双流CNN的制动蠕动颤振评价

李阳，靳畅*，李天舒，顾鼎元

作者信息 +

Evaluation of braking creeping flutter based on dual-stream CNN under fusion features

LI Yang, JIN Chang*, LI Tianshu, GU Dingyuan

Author information +

文章历史 +

摘要

针对车辆蠕动颤振主观评价方法效率低、耗时长、测试流程复杂的问题，研究了蠕动颤振信号的时序特征和时频域特征提取方法，将2D-CNN的空间处理能力与1D-CNN的时序处理能力相结合，提出一种融合特征下的双流卷积神经网络(DSCNN)的蠕动颤振评价方法。一条支路的输入为经过变分模态分解提取的时间序列特征，另一条支路的输入为经过快速傅里叶变换提取的图像特征，将一维时序特征与高维图像特征融合，训练模型进行评分。该方法通过融合不同模态的信息，充分捕捉蠕动颤振的局部波形特征和空间纹理特征。结果表明，融合两种特征的评分模型的八分类准确率达87.13%，验证了特征融合方法在蠕动颤振评价上的有效性。

Abstract

To solve the problems of low efficiency, long time consuming and complex test flow of subjective evaluation method for vehicle creep groan, the temporal features and time-frequency domain feature extraction method of creep groan signal are studied. A creep groan evaluation method based on dual-stream convolutional neural network (DSCNN) with fused features is proposed by combining the spatial processing ability of 2D-CNN with the temporal processing ability of 1D-CNN. One input is time series features extracted by variational modal decomposition, the other input is image features extracted by fast Fourier transform, the one-dimensional time series feature and the high-dimensional image feature are fused, and a training model is used for scoring. By fusing the information of different modes, the method can capture the local waveform features and spatial texture features of creep groan. The results show that the eight-classification accuracy of the scoring model is 87.13%, which verifies the effectiveness of the feature fusion method in creep groan evaluation.

导出引用

李阳, 靳畅, 李天舒, 顾鼎元. 融合特征下的双流CNN的制动蠕动颤振评价[J]. 振动与冲击, 2025, 44(1): 134-142

LI Yang, JIN Chang, LI Tianshu, GU Dingyuan. Evaluation of braking creeping flutter based on dual-stream CNN under fusion features[J]. Journal of Vibration and Shock, 2025, 44(1): 134-142

参考文献

[1] 张立军,张频捷,孟德建.汽车盘式制动器蠕动颤振试验与理论分析[J].汽车工程,2016,38(09):1132-1139.
ZHANG Lijun, ZHANG Pinjie, MENG Dejian.Tests and Theoretical Analysis of Creep Groan in Vehicle Disc Brake[J]. Automotive Engineering,2016,38(09):1132-1139.
[2] Bettella M, Harrison M, Sharp R S. Investigation of Auto-motive Creep Groan with a Disturbuted-source Excitation Technique[J]. Journal of Sound and Vibration, 2002, 255(3): 531-547.
[3] Abdelhamid M, Bray W.,"Braking Systems Creep Groan Noise: Detection and Evaluation," SAE Technical Paper 2009-01-2103, 2009
[4] Prezelj J , Murovec J , Huemer-Kals S ,et al.Identification of different manifestations of nonlinear stick–slip phenomena during creep groan braking noise by using the unsupervised learning algorithms k-means and self-organizing m-ap[J].Mechanical Systems and Signal Processing,Volume 166,2022,108349,ISSN 0888-3270
[5] 张立军,张兴,孟德建.汽车制动颤振瞬态特性与关键因素试验研究[J].机械工程学报,2018,54(24):118-128.
ZHANG Lijun, ZHANG Xing, MENG Dejian.Experiment study on transient characteristics and key factors of vehicle brake groan[J]. Journal of Mechanical Engineering, 2018,54(24):118-128.
[6] 张立军,黄健,孟德建, 等.基于道路试验的制动器蠕动颤振客观测量与主观评价[J].振动与冲击,2017,36(09):186-194.
ZHANG Lijun, HUANG Jian, MENG Dejian, et al. Objective measurement and subjective evaluation for brake creep groan based on road tests[J]. Journal of Vibration and Shock, 2017,36(09):186-194.
[7] 张立军,郑一兵,孟德建等.汽车制动器摩擦颤振综述[J].同济大学学报(自然科学版),2016,44(03):420-426+445.
ZHANG Lijun, ZHENG Yibing, MENG Dejian, et al. Literature survey of friction induced automotive brake groan[J]. Journal of Tongji University(Natural Science), 2016,44(03):420-426+445.
[8] Riefe M , Thompson J , Rogus M ,et al.SAE Low-Frequency Brake Noise Test Procedure[J]. 2010.
[9] 何成兵,车其祥,徐振华,等.基于参数自寻优变分模态分解的信号降噪方法[J].振动与冲击,2023,42(19):283-293.
HE Chengbing;CHE Qixiang;XU Zhenhua,et al. Signal denoising method based on parametric self-optimizing VMD[J]. Journal of Vibration and Shock, 2023,42(19):283-293.
[10] Zosso D , Dragomiretskiy K .Variational Mode Decom-position[J].IEEE Transactions on Signal Processing: A publication of the IEEE Signal Processing Society, 2014.
[11] 吴文轩,王志坚,张纪平等.基于峭度的VMD分解中k值的确定方法研究[J].机械传动,2018,42(08):153-157.
Wu Wenxuan;Wang Zhijian;Zhang Jiping,et al.Research of the Method of Determining k Value in VMD based on Kurtosis[J]. Journal of Mechanical Transmission, 2018, 42(08):153-157.
[12] 张露,理华,崔杰等.基于稀疏指标的优化变分模态分解方法[J].振动与冲击,2023,42(08):234-250.
ZHANG Lu;LI Hua;CUI Jie,et al. An optimal variational mode decomposition method based on sparse index[J]. Journal of Vibration and Shock,2023,42(08):234-250.
[13] 李宏斌,徐楚林,温周斌.基于短时傅里叶变换的异常音检测方法[J].声学技术,2014,33(02):145-149.
Li Hongbin; Xu Chulin; Wen Zhoubin. A STFT based method for detecting Rub & Buzz defects of loudspeaker and its application research[J]. Technical Acoustics, 2014,33(02):145-149.
[14] 姜洪权,贺帅,高建民,等.一种改进卷积神经网络模型的焊缝缺陷识别方法[J].机械工程学报,2020,56(08):235-242.
JIANG Hongquan;HE Shuai;GAO Jianmin,et al. An Improved Convolutional Neural Network for Weld Defect Recognition[J]. Journal of Mechanical Engineering, 2020,56(08):235-242.
[15] 宫文峰,陈辉,张泽辉,等.基于改进卷积神经网络的滚动轴承智能故障诊断研究[J].振动工程学报,2020,33(02):400-413.
GONG Wenfeng;CHEN Hui;ZHANG Zehui,et al. Intelligent fault diagnosis for rolling bearing based on improved convolutional neural network[J]. Journal of Vibration Engineering, 2020,33(2):400－413.
[16] 张训杰,张敏,李贤均.基于二维图像和CNN-BiGRU网络的滚动轴承故障模式识别[J].振动与冲击,2021,40(23):194-201207.
ZHANG Xunjie;ZHANG Min;LI Xianjun. Rolling bearing fault mode recognition based on 2D image and CNN-BiGRU[J]. Journal of Vibration and Shock, 2021,40(23):194-201207.
[17] Ye X , Li M .Press-fit process fault diagnosis using 1DCNN-LSTM method[J].Assembly Automation, 2022(3):42.
[18] L.J.P.V.D. Maaten;GE Hinton.Visualizing High-Dimensional Data using t-SNE[J].Journal of Machine Learning Research,2008,Vol.9