Health Indicator of Equipment Condition Monitoring Based on Adaptive Nonlinear State Estimation Under Compound Fault Degradation

doi:10.3901/JME.2025.03.119

Abstract

Abstract: Modern machinery and equipment have been in service for a long time under harsh and changeable working conditions and it is inevitable that multiple fault types will occur, resulting in a degradation process of compound fault conditions. Traditional health indicators usually focus on the condition monitoring of a single fault degradation situation and are not suitable for equipment condition monitoring in this situation. To solve the above problems, a new health indicator construction method based on adaptive nonlinear state estimation (ANSE) for equipment condition monitoring under coupled fault degradation is proposed, which includes two parts: initial fault transient detection and early fault accurate location. First, a singular value feature sequence is constructed based on Hilbert singular value decomposition (Hilbert-SVD) as input to the state monitoring model. Then, the nonlinear state estimation (NSE) reconstruction error characteristic is used to construct the HI to amplify the difference between the initial fault sample and the normal sample. After that, in order to self-adapt transient detection of initial faults, ANSE model is constructed by introducing peak overthreshold (POT) algorithm to realize dynamic update of health threshold. Finally, resonance based sparse signal decomposition and maximum correlation kurtosis deconvolution (RSSD-MCKD) feature extraction, using power spectrum analysis of fault characteristic frequency to achieve accurate fault location. The effectiveness and robustness of the proposed method are verified in the bearing data set of two coupled fault degradation cases and the engineering verification is carried out in one fan main bearing data set. By analyzing the test results and comparing the results, the proposed method can better detect the initial faults of the equipment and also characterize the entire degradation process, indicating that the proposed method has strong condition monitoring ability.

Key words: condition monitoring, health indicator, nonlinear state estimation, dynamic health threshold, fault characteristic frequency

CLC Number:

TH17
TN911

PEI Xuewu, LI Xinyu, GAO Liang, GAO Yiping, CHEN Zhimin. Health Indicator of Equipment Condition Monitoring Based on Adaptive Nonlinear State Estimation Under Compound Fault Degradation[J]. Journal of Mechanical Engineering, 2025, 61(3): 119-129.

References

[1] LI Xinyu，CAO Sican，GAO Liang，et al. A threshold-control generative adversarial network method for intelligent fault diagnosis[J]. Complex System Modeling and Simulation，2021，1(1):55-64.
[2] 沈天浩，丁康，黎杰，等. 图结构联合时序数据驱动的装备剩余使用寿命预测方法[J]. 机械工程学报，2023，59(12):183-194. SHEN Tianhao，DING Kang，LI Jie，et al. Residual service life prediction method driven by graph structure and time series data[J]. Journal of Mechanical Engineering，2019，59(12):183-194.
[3] LEI Yaguo，LI Naipeng，GUO Liang，et al. Machinery health prognostics:A systematic review from data acquisition to RUL prediction[J]. Mechanical Systems and Signal Processing，2018，104(1):799-834.
[4] CHEN Zhen，ZHOU Di，XIA Tangbin. Online unsupervised optimization framework for machine performance assessment based on distance metric learning[J]. Mechanical Systems and Signal Processing，2024，206(1):110883.
[5] KONG Weili，LI Hai. Remaining useful life prediction of rolling bearing under limited data based on adaptive time-series feature window and multi-step ahead strategy[J]. Applied Soft Computing，2022，129:109630.
[6] YANG Qichao，TANG Baoping，YANG Shilong，et al. An integrated network architecture for data repair and degradation trend prediction[J]. Mechanical Systems and Signal Processing，2023，200(1):110610.
[7] 肖扬，王庆锋，杨哲，等. 旋转机械突发不平衡故障早期预警及诊断方法研究[J]. 机械工程学报，2023，59(11):308-318. XIAO Yang，WANG Qingfeng，YANG Zhe，et al. Research on early warning and diagnosis method of sudden unbalance fault of rotating machinery[J]. Journal of Mechanical Engineering，2023，59(11):308-318.
[8] SHEN Yizhe，TANG Baoping，LI Biao，et al. Remaining useful life prediction of rolling bearing based on multi-head attention embedded Bi-LSTM network[J]. Measurement，2022，202:111803.
[9] LI Xiaomeng，WANG Yi，TANG Baoping，et al. Canonical correlation analysis of dimension reduced degradation feature space for machinery condition monitoring[J]. Mechanical Systems and Signal Processing，2023，182(1):109603.
[10] 王久健，杨绍普，刘永强，等. 一种基于空间卷积长短时记忆神经网络的轴承剩余寿命预测方法[J]. 机械工程学报，2021，57(21):88-95. WANG Jiujian，YANG Shaopu，LIU Yongqiang，et al. A bearing residual life prediction method based on spatial convolutional long and short time memory neural network[J]. Journal of Mechanical Engineering，2019，57(21):88-95.
[11] CHEN Dingliang，QIN Yi，WANG Yi，et al. Health indicator construction by quadratic function-based deep convolutional auto-encoder and its application into bearing RUL prediction[J]. ISA Transactions，2021，114:44-56.
[12] XIANG Sheng，ZHOU Jianghong，LUO Jun，et al. Cocktail lstm and its application into machine remaining useful life prediction[J]. IEEE/ASME Transactions on Mechatronics，2023，28(5):2425-2436.
[13] GUO Liang，YU Yaoxiang，DUAN Andongzhe，et al. An unsupervised feature learning based health indicator construction method for performance assessment of machines[J]. Mechanical Systems and Signal Processing，2022，167(15):108573.
[14] GONZÁLEZ-MUÑIZ A，DÍAZ I，CUADRADO A A，et al. Health indicator for machine condition monitoring built in the latent space of a deep autoencoder[J]. Reliability Engineering and System Safety，2022，224:108482.
[15] 申彦斌，张小丽，夏勇，等. Bi-LSTM神经网络用于轴承剩余使用寿命预测研究[J]. 振动工程学报，2021，34(2):411-420. SHEN Yanbin，ZHANG Xiaoli，XIA Yong，et al. Research on residual service life prediction of bearings using BI-LSTM neural network [J]. Journal of Vibration Engineering，2021，34(2):411-420.
[16] RAI A, KIM J M. A novel health indicator based on information theory features for assessing rotating machinery performance degradation[J]. IEEE Transactions on Instrumentation and Measurement，2020，69(9):6982-6994.
[17] ZHANG Guangyao，WANG Yi，LI Xiaomeng，et al. Health indicator based on signal probability distribution measures for machinery condition monitoring[J]. Mechanical Systems and Signal Processing，2023，1:110460.
[18] 李国正. 复杂工况下旋转机械的多源复合故障信号分离诊断方法研究[D]. 北京:北京化工大学，2021. LI Guozheng. Research on multi-source coupling fault signal separation diagnosis method of rotating machinery under complex working conditions[D]. Beijing:Beijing University of Chemical Technology，2021.
[19] YANG Shaohua，LIU Zhiliang，LU Guoliang，et al. Early change detection in dynamical bearing degradation process based on hierarchical graph model and adaptive inputs weighting fusion[J]. IEEE Transactions on Industrial Informatics，2021，17(5):3186-3196.
[20] 张龙，刘杨远，吴荣真，等. 基于MSET重构模型整体优化的轴承性能退化评估方法[J]. 振动与冲击，2023，42(16):251-261. ZHANG Long，LIU Yangyuan，WU Rongzhen，et al. Bearing performance degradation evaluation method based on MSET reconstruction model[J]. Journal of Vibration and Shock，2023，42(16):251-261.
[22] 邬世龙. 面向轴承剩余寿命预测的自适应退化阶段辨识方法[J]. 组合机床与自动化加工技术，2022(12):31-35. WU Shilong. Adaptive degradation stage identification method for bearing remaining life prediction[J]. Combined Machine Tool and Automatic Processing Technology，2022(12):31-35.
[23] 张守京，慎明俊，杨静雯，等. 改进的共振稀疏分解方法及其在滚动轴承复合故障诊断中的应用[J]. 中国机械工程，2022，33(14):1697-1706. ZHANG Shoujing，SHEN Mingjun，YANG Jingwen，et al. Improved resonance sparse decomposition method and its application in compound fault diagnosis of rolling bearing[J]. China Mechanical Engineering，2022，33(14):1697-1706.
[24] WANG Biao，LEI Yaguo，LI Naipeng，et al. A hybrid prognostics approach for estimating remaining useful life of rolling element bearings[J]. IEEE Transactions on Reliability，2020，69(1):401-412.
[25] QIU H，LEE J，LIN J，et al. Wavelet filter-based weak signature detection method and its application on rolling element bearing prognostics[J]. Journal of Sound and Vibration，2006，289(4-5):1066-1090.