Abstract:Aiming at the incomprehensibility of the data measured by a single sensor and the nonlinearity and non-stationarity of the fault signal of the centrifugal pump, a centrifugal pump cavitation-induced vibration fault based on the multi-fractal detrending fluctuation analysis (MFDFA) and BP neural network under multi-sensor data is proposed Analytical method. First, the MFDFA method is used to analyze the 8 types of measured signals of the centrifugal pump under 5 different working
conditions, and the multi-fractal spectrum characteristic parameters , , , and are extracted as the fault feature vector, and combined with the BP neural network for single sensor For signal fault diagnosis, the signal splicing with better recognition rate is selected to form a multi-sensor feature vector, and the multi-sensor centrifugal pump cavitation-induced vibration fault research is carried out. The results show that the characteristic parameters of the multi-fractal spectrum extracted by the MFDFA method can accurately reflect the operation status of the pump, among which parameters such as , and have a better effect on fault classification; signals such as pump vibration, torque and motor vibration reflect the nature of the fault more. Accurate; the accuracy of the multi-sensor fault diagnosis model formed on this basis is improved by more than 13% than that of the single sensor, which provides a new method for the state recognition of the centrifugal pump with different degrees of cavitation fault.
Keywords: centrifugal pump; multi-sensor data; fault diagnosis; multi-fractal; detrended fluctuation analysis; BP neural network
梁兴,罗远兴,邓飞,高刚刚,曹寒问. 多传感器数据下基于MFDFA-BP的离心泵空化致振故障分析[J]. 振动与冲击, 2022, 41(17): 238-243.
LIANG Xing, LUO Yuanxing, DENG Fei, GAO Ganggang, CAO Hanwen. Cavitation induced vibration fault analysis of centrifugal pump based on MFDFA-BP under multi-sensor data. JOURNAL OF VIBRATION AND SHOCK, 2022, 41(17): 238-243.
[1] Li W , Zhu Z , Jiang F , et al. Fault diagnosis of rotating machinery with a novel statistical feature extraction and evaluation method[J]. Mechanical Systems and Signal Processing, 2015, s 50–51(jan.):414–426.
[2] Ming Z , Jiang Z , Feng K . Research on variational mode decomposition in rolling bearings fault diagnosis of the multistage centrifugal pump[J]. Mechanical Systems and Signal Processing, 2017, 93(Sep.):460-493.
[3] Sun H , Si Q , Chen N , et al. HHT-based feature extraction of pump operation instability under cavitation conditions through motor current signal analysis[J]. Mechanical Systems and Signal Processing, 2020, 139:106613.
[4] Morsy M E , Abouel-Seoud S A , Rabeih E A . Gearbox Damage Diagnosis Using Wavelet Transform Technique[J]. International Journal of Acoustics & Vibration, 2011, 16(4).
[5] Kantelhardt J W , Zschiegner S A , Koscielny-Bunde E,et al. Multifractal detrended fluctuation analysis of nonstationary time series[J]. Physica A: Statistical Mechanics and its Applications, 2002,316(1):87-114.
[6] Logan D B , Mathew J . Using the correlation dimension for vibration fault diagnosis of rolling element bearings—II. Selection of experimental parameters[J]. Mechanical Systems and Signal Processing, 1996, 10(3):251-264.
[7] Purkait P , Chakravorti S . Impulse fault classification in transformers by fractal analysis[J]. Dielectrics & Electrical Insulation IEEE Transactions on, 2003, 10(1):109-116.
[8] 李兆飞, 柴毅, 李华锋. 多重分形去趋势波动分析的振动信号故障诊断[J]. 华中科技大学学报(自然科学版), 2012, 40(012):5-9.
Li Zhaofei, Chai Yi, Li Huafeng. Vibration signal troubleshooting for multi-fractal de-trend fluctuation analysis. Journal of Huazhong University of Science and Technology (Natural Science Edition), 2012, 40 (012): 5-9. (in Chinese)
[9] 裴峻峰, 严安, 彭剑,等. 声信号的MFDFA和SFLA-SVM算法的往复泵故障诊断[J]. 机械设计与制造, 2020, 000(004):199-203,207.
Yan Junfeng, Yan An, Peng Jian, etc. Reciprocating pump fault diagnosis of the MFDFA and SFLA-SVM algorithms for acoustic signals. Mechanical Design and Manufacturing, 2020, 000 (004): 199-203, 207. (in Chinese)
[10] 赵建光,李红波,曾繁景,李铁峰.模拟电路的马氏距离故障诊断方法研究[J].电子测量技术,2012,35(03):128-131.
Zhao Jianguang, Li Hongbo, Zeng Fanjing, Li Tiefeng. Study on Mars distance fault diagnosis method of analog circuit, 2012, 35 (03): 128-131. (in Chinese)
[11] Talebinejad M , Chan A D C , Miri A . Fatigue estimation using a novel multi-fractal detrended fluctuation analysis-based approach[J]. Journal of Electromyography & Kinesiology, 2010, 20(3):433-439.
[12] Stanley H E , Meakin P . Multifractal phenomena in physics and chemistry[J]. Nature, 1988, 335(6189):405-409.
[13] Moktadir Z , Kraft M , Wensink H . Multifractal properties of Pyrex and silicon surfaces blasted with sharp particles[J]. Physica A: Statistical Mechanics and its Applications, 2008, 387(8-9):2083-2090.
[14] Rumelhart D E , Mcclelland J L , Group C P . Parallel distributed processing: explorations in the microstructure of cognition, vol. 2: psychological and biological models[J]. Language, 1986, 63(4):45 - 76.
[15] 郭鹏程, 孙龙刚, 李辉,等. 基于多重分形谱和改进BP神经网络的水电机组振动故障诊断研究[J]. 水力发电学报, 2014, 33(3):299-305.
Guo Pengcheng, Sun Longgang, Li Hui, etc. Research on vibration fault diagnosis of hydropower units based on multiple fractal spectrometry and improvement of BP neural network. Journal of Hydroelectric Power, 2014, 33 (3): 299-305. (in Chinese)
[16] 张建军, 张利, 穆海芳,等. 基于改进粒子群优化BP网络的发动机故障诊断方法[J]. 农业机械学报, 2011(01):198-203.
Zhang Jianjun, Zhang Li, Mu Haifang, etc. Engine fault diagnosis method based on improving particle group optimization BP network. Journal of Agricultural Machinery, 2011 (01): 198-203. (in Chinese)
[17] 鲁娟娟,陈红.BP神经网络的研究进展[J].控制工程,2006(05):449-451+456.
Lu Wei, Chen Hong. Advances in the study of BP neural networks, Control Engineering, 2006 (05): 449-451-456. (in Chinese)
[18] 卿彪, 任建军, 宋晓飞. 基于EMD的离心泵空化初生的研究[J]. 水力发电学报, 2016, v.35;No.166(05):102-109.
Qin Biao, Ren Jianjun, Song Xiaofei.Research on cavitation initiation of centrifugal pump based on EMD [J]. Journal of Hydroelectric Engineering, 2016, v.35;No.166(05):102-109. (in Chinese)