基于补偿距离评估和一维卷积神经网络的离心泵故障快速智能识别方法

PDF(1890 KB)

振动与冲击 ›› 2021, Vol. 40 ›› Issue (10) : 41-49.

论文

焦瀚晖1，胡明辉1，江志农1，冯坤2

作者信息 +

Fast and intelligent identification method for faults of a centrifugal pump based on the compensation distance evaluation and one-dimensional convolution neural network

JIAO Hanhui1，HU Minghui1，JIANG Zhinong1，FENG Kun2

Author information +

文章历史 +

摘要

一维卷积神经网络可适用于振动等一维信号的识别与分类，但将其直接应用于机械故障诊断时小样本训练条件下的识别准确率与识别速度是其亟需解决的问题。针对上述问题，提出一种基于补偿距离评估和一维卷积神经网络的离心泵故障快速智能识别方法。基于离心泵振动分析与故障诊断理论，通过提取时域、频域、能量及熵特征来构造混合域全特征集，充分挖掘训练样本中的故障信息，提高单个训练样本的利用率，使故障识别模型具备小样本训练的能力；通过补偿距离评估方法对全特征集进行降维优化，在有效保留故障特征信息的同时显著降低特征维度，使特征构造及故障识别模型具备快速计算的能力；通过训练样本的降维后特征进行一维卷积神经网络的训练，进而构建故障智能识别模型，保存模型并将其用于离心泵故障分析。经某石化离心泵的抽空和滚动轴承损伤两个故障案例验证，该方法在小样本训练条件下识别准确率达到98%以上，单组数据识别时间小于3 s，可满足工程中离心泵故障实时智能识别的需求。

Abstract

1D convolution neural network (1DCNN) can be applied to the identification and classification of one-dimensional signals, such as vibration signal, etc.But, the recognition accuracy and recognition speed with small sample training are crucial problems to be solved, when it is directly applied to mechanical fault diagnosis.To solve these problems, a fast and intelligent fault identification method based on the compensation distance evaluation technique (CDET) and 1DCNN was proposed.First, based on the centrifugal pump vibration analysis and fault diagnosis theory, a full-featured set in mixed domains was constructed by extracting time-domain, frequency-domain, energy, and entropy features.The full-featured set can fully mine the fault information in the training samples, improve the utilization rate of a single training sample and let the fault recognition model be able to train by small sample.Then, the full-featured set was reduced in dimension and optimized through the CDET, while effectively retaining the fault feature information, so that the feature construction and fault identification model was provided with the ability of quick calculation.Finally, the DCNN was trained by the reduced dimension features of the training samples, and then the fault intelligent identification model was constructed for the fault analysis of the centrifugal pump.Through two fault cases of a petrochemical centrifugal pump, i.e., evacuation and rolling bearing damage, it is proved that the identification accuracy of the method is more than 98% under the condition of small sample training, and the recognition time of a single group of data is less than 3 s, which can meet the needs of real-time intelligent identification of faults of the centrifugal pump in engineering.

导出引用

焦瀚晖1，胡明辉1，江志农1，冯坤2. 基于补偿距离评估和一维卷积神经网络的离心泵故障快速智能识别方法[J]. 振动与冲击, 2021, 40(10): 41-49

JIAO Hanhui1，HU Minghui1，JIANG Zhinong1，FENG Kun2. Fast and intelligent identification method for faults of a centrifugal pump based on the compensation distance evaluation and one-dimensional convolution neural network[J]. Journal of Vibration and Shock, 2021, 40(10): 41-49

参考文献

［1］姚家伦,李龙森.机床齿轮箱智能振动故障诊断专家系统［J］.振动与冲击, 1988,7(3): 73-78.
YAO Jialun, LI Longsen.Intelligent vibration fault diagnosis expert system for machine tool gearbox［J］.Journal of Vibration and Shock, 1988,7(3): 73-78.

［2］夏松波, 张嘉钟, 徐世昌, 等.旋转机械故障诊断技术的现状与展望［J］.振动与冲击, 1997,16(2): 5-9.
XIA Songbo, ZHANG Jiazhong, XU Shichang, et al.Present situation and prospect of fault diagnosis technology for rotating machinery［J］.Journal of Vibration and Shock, 1997,16(2): 5-9.

［3］王青华, 杨天海, 沈润杰, 等.抽水蓄能机组振动故障诊断专家系统［J］.振动与冲击, 2012,31(7): 158-161.
WANG Qinghua,YANG Tianhai, SHEN Runjie, et al.Fault caused by vibration diagnosis expert system for a pump storage group［J］.Journal of Vibration and Shock, 2012,31(7): 158-161.

［4］刘星辰, 周奇才, 赵炯, 等.一维卷积神经网络实时抗噪故障诊断算法［J］.哈尔滨工业大学学报, 2019,51(7): 89-95.
LIU Xingchen, ZHOU Qicai, ZHAO Jiong, et al.Real-time and anti-noise fault diagnosis algorithm based on 1D convolutional neural network［J］.Journal of Harbin Institute of Technology, 2019,51(7): 89-95.

［5］ZHANG W, PENG G L, LI C H.Bearings fault diagnosis based on convolutional neural networks with 2D representation of vibration signals as input［C］//MATEC Web of Conferences.Shanghai: MATEC, 2017.

［6］LONG W, LI X Y, LIANG G, et al.A new convolutional neural network-based data-driven fault diagnosis method［J］.IEEE Transactions on Industrial Electronics, 2018,65(7): 5990-5998.

［7］李恒, 张氢, 秦仙蓉, 等.基于短时傅里叶变换和卷积神经网络的轴承故障诊断方法［J］.振动与冲击, 2018,37(19): 132-139.
LI Heng, ZHANG Qing, QIN Xianrong, et al.Fault diagnosis method for rolling bearings based on short-time Fourier transform and convolution neural network［J］.Journal of Vibration and Shock, 2018,37(19): 132-139.

［8］王丽华, 谢阳阳, 周子贤, 等.基于卷积神经网络的异步电机故障诊断［J］.振动、测试与诊断, 2017,37(6): 1208-1215.
WANG Lihua, XIE Yangyang, ZHOU Zixian, et al.Fault diagnosis of asynchronous motor based on convolutional neural network［J］.Journal of Vibration, Measurement & Diagnosis, 2017,37(6): 1208-1215.

［9］曾雪琼.基于卷积神经网络的变速器故障分类识别研究［D］.广州：华南理工大学, 2016.

［10］SUN W, YAO B, ZENG N Y, et al.An intelligent gear fault diagnosis methodology using a complex wavelet enhanced convolutional neural network［J］.Materials, 2017,10(7): 790.

［11］JIA F, LEI Y G, LU N, et al.Deep normalized convolutional neural network for imbalanced fault classification of machinery and its understanding via visualization［J］.Mechanical Systems and Signal Processing, 2018,110: 349-367.

［12］胡晓依, 荆云建, 宋志坤, 等.基于CNN-SVM的深度卷积神经网络轴承故障识别研究［J］.振动与冲击, 2019,38(18): 173-178.
HU Xiaoyi, JING Yunjian, SONG Zhikun, et al.Bearing fault identification by using deep convolution neural networks based on CNN-SVM［J］.Journal of Vibration and Shock, 2019,38(18): 173-178.

［13］马慧.基于状态的滚动轴承寿命预测与维修计划优化研究［D］.北京：北京交通大学, 2017.

［14］雷亚国, 何正嘉,訾艳阳.基于混合智能新模型的故障诊断［J］.机械工程学报, 2008,44(7): 112-117.
LEI Yaguo, HE Zhengjia, ZI Yanyang.Fault diagnosis based on novel hybrid intelligent model［J］.Journal of Mechanical Engineering, 2008,44(7): 112-117.

［15］王鑫.基于小波变换的机械轴承磨损故障特征提取方法研究［D］.成都：西南交通大学, 2016.

［16］HU X, HE Z J, ZHANG Z S, et al.Fault diagnosis of rotating machinery based on improved wavelet package transform and SVMs ensemble［J］.Mechanical Systems & Signal Processing, 2007,21(2): 688-705.

［17］LEI Y G, HE Z J, ZI Y Y, et al.New clustering algorithm-based fault diagnosis using compensation distance evaluation technique［J］.Mechanical Systems & Signal Processing, 2008,22(2): 419-435.

［18］王宏超, 陈进,董广明.基于补偿距离评估-小波核PCA的滚动轴承故障诊断［J］.振动与冲击, 2013,32(18): 87-90.
WANG Hongchao, CHEN Jin, DONG Guangming.Fault diagnosis of rolling bearing based on compensation distance evaluation technique-wavelet kernel principal component analysis［J］.Journal of Vibration and Shock, 2013,32(18): 87-90.

［19］陈雯柏.人工神经网络原理与实践［M］.西安: 西安电子科技大学出版社,2015.

［20］SCHLKOPF B, PLATT J, HOFMANN T.Greedy layer-wise training of deep networks［J］.Advances in Neural Information Processing Systems, 2007,19: 153-160.

［21］吴春志, 江鹏程, 冯辅周, 等.基于一维卷积神经网络的齿轮箱故障诊断［J］.振动与冲击, 2018,37(22): 51-56.
WU Chunzhi, JIANG Pengcheng, FENG Fuzhou, et al.Faults diagnosis method for gearboxes based on a 1D convolutional neural network［J］.Journal of Vibration and Shock, 2018,37(22): 51-56.

［22］林克正, 白婧轩, 李昊天, 等.深度学习下融合不同模型的小样本表情识别［J］.计算机科学与探索, 2020,14(3): 482-492.
LIN Kezheng, BAI Jingxuan, LI Haotian, et al.Facial expression recognition with small samples fused with different models under deep learning［J］.Journal of Frontiers of Computer Science and Technology, 2020,14(3): 482-492.