声振信号融合的堆垛机故障诊断

doi:10.3778/j.issn.1002-8331.2409-0264

摘要/Abstract

摘要： 工业堆垛机的传动部件故障信号不明显，特征识别效率低，且传统网络模型参数量大，限制了故障诊断的效率与精度。针对上述问题提出一种声振信号融合的堆垛机故障诊断方法。针对不同信号特征的有效提取与优化问题，提出变分模态分解（variational mode decomposition，VMD）和改进格拉姆角差场（improved Gramian angular difference field，IGADF）特征工程方法，对信号进行降噪、低频模态重构、优化的图像编码处理。设计多模态特征融合网络，融合八种模态的图像特征，以丰富的模态信息提升模型决策准确性和鲁棒性。针对传统诊断模型的效率与精度问题。对ResNet34网络改进第二卷积层和激活函数，加入注意力机制并调整残差块数量，使其更适用于数据集，降低网络参数量和计算复杂度。实验结果表明，改进方法在自采集数据集和公开数据集MFPT上的准确率分别达98.5%和98.1%，在参数量较低的条件下展现出优异的故障诊断性能和良好的泛化能力。

关键词: 堆垛机故障诊断, 声振信号融合, 改进格拉姆角差场, 多模态特征融合, 改进ResNet34

Abstract: The fault signal of the transmission component of the industrial stacker is not obvious, the feature recognition efficiency is low, and the traditional network model has a large number of parameters, which limits the efficiency and accuracy of fault diagnosis. To address the above problems, a stacker fault diagnosis method based on acoustic and vibration signal fusion is proposed. In order to effectively extract and optimize the features of different signals, the variational mode decomposition(VMD) and improved Gramian angular difference field (IGADF) feature engineering methods are proposed to reduce noise, reconstruct low-frequency modes, and optimize image coding for the signal. Then a multimodal feature fusion network is designed to fuse the image features of eight modes, and the model decision accuracy and robustness are improved with rich modal information. Finally, the efficiency and accuracy of the traditional diagnosis model are addressed. The second convolutional layer and activation function of the ResNet34 network are improved, the attention mechanism is added, and the number of residual blocks is adjusted to make it more suitable for the data set and reduce the network parameters and computational complexity. Experimental results show that the improved method has an accuracy of 98.5% and 98.1% on the self-collected dataset and the public dataset MFPT, respectively, showing excellent fault diagnosis performance and good generalization ability under the condition of low parameter number.

Key words: stacker fault diagnosis, acoustic and vibration signal fusion, improved Gramian angular difference field, multimodal feature fusion, improved ResNet34

谢雷, 孟文, 孟祥印, 李杨, 李世初. 声振信号融合的堆垛机故障诊断[J]. 计算机工程与应用, 2025, 61(24): 346-359.

XIE Lei, MENG Wen, MENG Xiangyin, LI Yang, LI Shichu. Vibration and Sound Signal Fusion for Stacker Crane Fault Diagnosis[J]. Computer Engineering and Applications, 2025, 61(24): 346-359.

参考文献

[1] FU Y, CAO H R, CHEN X F, et al. Task-incremental broad lea-rning system for multi-component intelligent fault diagnosis of machinery[J]. Knowledge-Based Systems, 2022, 246: 108730.
[2] ZHANG K Y, CHEN J L, HE S L, et al. Differentiable neural architecture search augmented with pruning and multi-objective optimization for time-efficient intelligent fault diagnosis of machinery[J]. Mechanical Systems and Signal Processing, 2021, 158: 107773.
[3] 那峙雄, 孙涛, 来广志, 等. 多尺度特征融合的光伏电站故障诊断[J]. 计算机工程与应用, 2022, 58(10): 300-308.
NA Z X, SUN T, LAI G Z, et al. Fault diagnosis for photovoltaic power station by multi-scale features fusion[J]. Computer Engineering and Applications, 2022, 58(10): 300-308.
[4] ZHAO X L, JIA M P. A novel unsupervised deep learning network for intelligent fault diagnosis of rotating machinery[J]. Structural Health Monitoring, 2020, 19(6): 1745-1763.
[5] 李莎莎, 石颉. 面向感应电机故障诊断的深度学习方法研究[J]. 计算机工程与应用, 2024, 60(14): 329-336.
LI S S, SHI J. Research on deep learning method for induction motor fault diagnosis[J]. Computer Engineering and Applications, 2024, 60(14): 329-336.
[6] 史丽晨, 张鹏, 王海涛, 等. 基于GASF与MSCAM-DenseNet的小样本齿轮故障诊断方法[J]. 计算机集成制造系统, 2025, 31(8): 3033-3045.
SHI L C, ZHANG P, WANG H T, et al. Small-sample gear fault diagnosis method based on GASF and MSCAM-DenseNet[J]. Computer Integrated Manufacturing Systems, 2025, 31(8): 3033-3045.
[7] 周玉蓉, 张巧灵, 于广增, 等. 基于声信号的工业设备故障诊断研究综述[J]. 计算机工程与应用, 2023, 59(7): 51-63.
ZHOU Y R, ZHANG Q L, YU G Z, et al. Review of acoustic signal-based industrial equipment fault diagnosis[J]. Computer Engineering and Applications, 2023, 59(7): 51-63.
[8] HUANG D R, KE L Y, MI B, et al. A new incipient fault diagnosis method combining improved RLS and LMD algorithm for rolling bearings with strong background noise[J]. IEEE Access, 2018, 6: 26001-26010.
[9] PENG Z K, TSE P W, CHU F L. A comparison study of imp-roved Hilbert Huang transform and wavelet transform: application to fault diagnosis for rolling bearing[J]. Mechanical Systems and Signal Processing, 2005, 19(5): 974-988.
[10] SHANG X Q, HUANG T L, CHEN H P, et al. Recursive variational mode decomposition enhanced by orthogonalization algorithm for accurate structural modal identification[J]. Mechanical Systems and Signal Processing, 2023, 197: 110358.
[11] XI C P, LIU R Q. Detection of small floating target on sea surface based on gramian angular field and improved EfficientNet[J]. Remote Sensing, 2022, 14(17): 4364.
[12] NASCIMENTO E G S, DE MELO T A C, MOREIRA D M. A transformer-based deep neural network with wavelet transform for forecasting wind speed and wind energy[J]. Energy, 2023, 278: 127678.
[13] HE K C, XU Y W, WANG Y, et al. Intelligent diagnosis of rolling bearings fault based on multisignal fusion and MTF-ResNet[J]. Sensors, 2023, 23(14): 6281.
[14] STEFENON S F, SEMAN L O, AQUINO L S, et al. Wavelet-Seq2Seq-LSTM with attention for time series forecasting of level of dams in hydroelectric power plants[J]. Energy, 2023, 274: 127350.
[15] GONG Q Y, PENG K, GAO Q, et al. Series arc fault identification method based on wavelet transform and feature values decomposition fusion DNN[J]. Electric Power Systems Research, 2023, 221: 109391.
[16] CAI Y Y, YAO Z K, CHENG X, et al. Deep metric learning framework combined with Gramian angular difference field image generation for Raman spectra classification based on a handheld Raman spectrometer[J]. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 2023, 303: 123085.
[17] XIONG L J, HE M, HU C, et al. Image presentation and effective classification of odor intensity levels using multi-channel electronic nose technology combined with GASF and CNN[J]. Sensors and Actuators B: Chemical, 2023, 395: 134492.
[18] CHANG M, YAO D C, YANG J W. Intelligent fault diagnosis of rolling bearings using efficient and lightweight ResNet networks based on an attention mechanism (September 2022)[J]. IEEE Sensors Journal, 2023, 23(9): 9136-9145.
[19] YIN X L, MOU Z L, WANG Y Q. Fault diagnosis of wind turbine gearbox based on multiscale residual features and ECA-stacked ResNet[J]. IEEE Sensors Journal, 2023, 23(7): 7320-7333.
[20] 万昕, 刘坤, 崔昌浩. 一种基于Sobel梯度的直方图均衡算法及其在红外图像上的应用[J]. 红外技术, 2024, 46(4): 452-459.
WAN X, LIU K, CUI C H. Histogram equalization algorithm based on sobel gradient and its application on infrared images[J]. Infrared Technology, 2024, 46(4): 452-459.
[21] 徐剑, 丁晓青, 王生进, 等. 一种融合局部纹理和颜色信息的背景减除方法[J]. 自动化学报, 2009, 35(9): 1145-1150.
XU J, DING X Q, WANG S J, et al. Background subtraction based on a combination of local texture and color[J]. Acta Automatica Sinica, 2009, 35(9): 1145-1150.
[22] 卫少萌, 张江民, 石慧, 等. 融合多传感器数据的系统相对密度高维核估计剩余寿命预测[J]. 计算机集成制造系统, 2024, 30(12): 4493-4507.
WEI S M, ZHANG J M, SHI H, et al. Remaining useful life prediction of relative density high-dimensional kernel estimation for systems based on multi-sensor data fusion[J]. Computer Integrated Manufacturing Systems, 2024, 30(12): 4493-4507.
[23] 苏兆婧, 余隋怀, 初建杰, 等. 面向云服务平台的产品感性评价及标注模型[J]. 计算机集成制造系统, 2021, 27(3): 868-877.
SU Z J, YU S H, CHU J J, et al. Evaluation and annotation model of product Kansei attributes on cloud service platform[J]. Computer Integrated Manufacturing Systems, 2021, 27(3): 868-877.
[24] 万鹏, 赵竣威, 朱明, 等. 基于改进Res Net50模型的大宗淡水鱼种类识别方法[J]. 农业工程学报, 2021, 37(12): 159-168.
WAN P, ZHAO J W, ZHU M, et al. Freshwater fish species identification method based on improved Res Net50 model[J]. Transactions of the Chinese Society of Agricultural Engineering, 2021, 37(12): 159-168.
[25] 张宏鸣, 沈寅威, 阳光, 等. 融合注意力机制与多尺度信息的葡萄种植区变化检测[J]. 农业机械学报, 2024, 55(5): 196-206.
ZHANG H M, SHEN Y W, YANG G, et al. Change detection of grape growing areas based on integrating attention mechanism and multiscale information[J]. Transactions of the Chinese Society for Agricultural Machinery, 2024, 55(5): 196-206.
[26] LI H L, LI J, WEI H B, et al. Slim-neck by GSConv: a lightweight-design for real-time detector architectures[J]. arXiv: 2206. 02424, 2022.
[27] 刘祥, 田敏, 梁金艳. 基于RCH-UNet的新疆密植棉花图像快速分割及产量预测[J]. 农业工程学报, 2024, 40(7): 230-239.
LIU X, TIAN M, LIANG J Y. Prediction of cotton yield densely planted in Xinjiang of China using RCH-UNet model[J]. Transactions of the Chinese Society of Agricultural Engineering, 2024, 40(7): 230-239.
[28] DENG Y J, HOU Y L, YAN J T, et al. ELU-Net: an efficient and lightweight U-Net for medical image segmentation[J]. IEEE Access, 2022, 10: 35932-35941.
[29] DOU S Q, WANG L, FAN D L, et al. Classification of citrus huanglongbing degree based on CBAM-MobileNetV2 and transfer learning[J]. Sensors, 2023, 23(12): 5587.
[30] ZHENG Q H, SAPONARA S, TIAN X Y, et al. A real-time constellation image classification method of wireless communication signals based on the lightweight network MobileViT[J]. Cognitive Neurodynamics, 2024, 18(2): 659-671.
[31] ZHENG Q H, TIAN X Y, YU Z G, et al. MobileRaT: a lightweight radio transformer method for automatic modulation classification in drone communication systems[J]. Drones, 2023, 7(10): 596.