Observer-biased bearing condition monitoring: From fault detection to multi-fault classification

Abstract

Bearings are simultaneously a fundamental component and one of the principal causes of failure in rotary machinery. The work focuses on the employment of fuzzy clustering for bearing condition monitoring, i.e., fault detection and classification. The output of a clustering algorithm is a data partition (a set of clusters) which is merely a hypothesis on the structure of the data. This hypothesis requires validation by domain experts. In general, clustering algorithms allow a limited usage of domain knowledge on the cluster formation process. In this study, a novel method allowing for interactive clustering in bearing fault diagnosis is proposed. The method resorts to shrinkage to generalize an otherwise unbiased clustering algorithm into a biased one. In this way, the method provides a natural and intuitive way to control the cluster formation process, allowing for the employment of domain knowledge to guiding it. The domain expert can select a desirable level of granularity ranging from fault detection to classification of a variable number of faults and can select a specific region of the feature space for detailed analysis. Moreover, experimental results under realistic conditions show that the adopted algorithm outperforms the corresponding unbiased algorithm (fuzzy c-means) which is being widely used in this type of problems.

Harmônicos de Referência Observados por Sistemas Inteligentes na Identificação de Falhas Elétricas em Motores de Indução

“…Várias técnicas têm sido pesquisadas e demonstradas para a identificação e classificação de anomalias de operação de forma preditiva, como demonstrado nos trabalhos de Mabrouk and Zouzou (2015); Li et al (2016), e Liu et al (2018). A escolha da técnica a ser empregada considera a dinâmica do processo, a operação do equipamento, o tipo de acionamento ou controle e os critérios adicionais diretamente relacionadosàs respostas dos métodos, Lakehal and Ramdane (2017).…”

“…A diversidade de métodos diagnósticos pode ser vista em Thomson and Fenger (2001) com as ferramentas de processamento de sinais como a Transformada Rápida de Fourier (FFT) aplicada em sinais de corrente elétrica, ındices de vibração e ruídos acústicos. Considerando a empregabilidade maior da Análise dos Sinais das Correntes do estator do Motor (MCSA), as mudanças de padrões de amplitude no espectro do sinal ou nos componentes espectrais da FFT permitem caracterizar e acompanhar a evolução das falhas específicas desde o seu início até inoperância da máquina, Li et al (2016); de Jesus Romero-Troncoso (2017); Mata-Castrejón et al (2015).…”

Clustering algorithm using rough set theory for unsupervised feature selection

“…Each raw vibration signal is processed to get statistical parameters in three domains, i.e, time, frequency, time-frequency. 1622 features are obtained, and after the correlated features were removed the database ended up with 478 features [17].…”

Bearing Fault Feature Selection Method Based on Weighted Multidimensional Feature Fusion

“…With the rapid development of data mining, computer technology and artificial intelligence [15], data-driven fault diagnosis methods have increasingly shown their strong applicability, and often use a combination of three methods. Based on the redundant second generation wavelet packet transform (RSGWPT), Liu et al [16] extracted 56 features of the vibration signal and input support vector machine (SVM) for fault identification; Tian et al [17] selected permutation entropy (PE) as the fault feature, and proposed a manifold-based dynamic time warping method for fault diagnosis; Li et al [18] selected 1634 features and classified the bearing faults using the method of fuzzy C-means with a variable focal point (FCMFP); In [19], composite multiscale fuzzy entropy (CMFE) was selected as the feature to train the ensemble support vector machine (ESVM) for fault diagnosis of the rolling element bearings; In [20], the energy entropy of the intrinsic mode function (IMFs) of the bearing vibration signal is extracted, and combined with probabilistic neural network (PNN) and simplified fuzzy adaptive resonance theory map (SFAM) for online bearing fault diagnosis; In [21], the hierarchical symbol dynamic entropy (HSDE) is used as a sensitive feature input binary tree support vector machine (BT-SVM) to effectively identify the fault of the bearing. Most of these tasks use statistic and signal analysis to extract the features of vibration signals and to diagnose faults based on artificial intelligence.…”