Deep learning and its applications to machine health monitoring

This paper investigates diagnostic techniques for electrical machines with special reference to induction machines and to papers published in the last ten years. A comprehensive list of references is reported and examined, and research activities classified into four main topics: 1) electrical faults; 2) mechanical faults; 3) signal processing for analysis and monitoring; and 4) artificial intelligence and decision-making techniques.

Advances in Diagnostic Techniques for Induction Machines

Ordinary differential equations of the first order linear differential equations complex numbers and linear algebra simultaneous linear differential equations numerical methods the descriptive theory of nonlinear differential equations mechanical systems and electric circuits Fourier series Fourier integrals and Fourier transforms the Laplace transformation partial differential equations Bessel functions and Legendre polynomials applications and further properties of matrices vector analysis the calculus of variations analytic functions of a complex variable infinite series in the complex plane the theory of residues conformal mapping.

Advanced Engineering Mathematics. By C. R. Wylie. Pp. xiv, 813. 1966. (McGraw-Hill.)

A sparse auto-encoder-based deep neural network approach for induction motor faults classification

In modern manufacturing systems and industries, more and more research efforts have been made in developing effective machine health monitoring systems. Among various machine health monitoring approaches, data-driven methods are gaining in popularity due to the development of advanced sensing and data analytic techniques. However, considering the noise, varying length and irregular sampling behind sensory data, this kind of sequential data cannot be fed into classiﬁcation and regression models directly. Therefore, previous work focuses on feature extraction/fusion methods requiring expensive human labor and high quality expert knowledge. With the development of deep learning methods in the last few years, which redeﬁne representation learning from raw data, a deep neural network structure named Convolutional Bi-directional Long Short-Term Memory networks (CBLSTM) has been designed here to address raw sensory data. CBLSTM ﬁrstly uses CNN to extract local features that are robust and informative from the sequential input. Then, bi-directional LSTM is introduced to encode temporal information. Long Short-Term Memory networks(LSTMs) are able to capture long-term dependencies and model sequential data, and the bi-directional structure enables the capture of past and future contexts. Stacked, fully-connected layers and the linear regression layer are built on top of bi-directional LSTMs to predict the target value. Here, a real-life tool wear test is introduced, and our proposed CBLSTM is able to predict the actual tool wear based on raw sensory data. The experimental results have shown that our model is able to outperform several state-of-the-art baseline methods.

/pdf/learning-to-monitor-machine-health-with-convolutional-bi-2vok1ch88q.pdf

Learning to Monitor Machine Health with Convolutional Bi-Directional LSTM Networks.

Condition monitoring is desirable for increasing machinery availability, reducing consequential damage, and improving operational efficiency. Model-based methods are efficient monitoring systems for providing warning and predicting certain faults at early stages. However, the conventional methods must work with explicit motor models, and cannot be applied effectively for vibration signal diagnosis due to their nonadaptation and the random nature of vibration signal. In this paper, an analytical redundancy method using neural network modeling of the induction motor in vibration spectra is proposed for machine fault detection and diagnosis. The short-time Fourier transform is used to process the quasi-steady vibration signals to continuous spectra for the neural network model training. The faults are detected from changes in the expectation of vibration spectra modeling error. The effectiveness of the proposed method is demonstrated through experimental results, and it is shown that a robust and automatic induction machine condition monitoring system has been produced

Induction Machine Condition Monitoring Using Neural Network Modeling

Fault detection is desirable for increasing machinery availability, reducing consequential damage, and improving operational efficiency. Many of these faulty situations in three-phase induction motors originate from an electrical source. Vibration signal analysis is found to be sensitive to electrical faults. However, conventional methods require detailed information on motor design characteristics and cannot be applied effectively to vibration diagnosis because of their nonadaptability and the random nature of the vibration signals. This paper presents the development of an online electrical fault detection system that uses neural network modeling of induction motor in vibration spectra. The short-time Fourier transform is used to process the quasi-steady vibration signals for continuous spectra so that the neural network model can be trained. The electrical faults are detected from changes in the expectation of modeling errors. Experimental observations show that a robust and automatic electrical fault detection system is produced whose effectiveness is demonstrated while minimizing the triggering of false alarms due to power supply imbalance.

Vibration signal analysis for electrical fault detection of induction machine using neural networks

This paper presents the development of an online electrical fault detection system that uses neural network (NN) modeling of induction motor in vibration spectra. The short-time Fourier transform (STFT) is used to process the quasi-steady vibration signals for continuous spectra so that the NN model can be trained. The electrical faults are detected from changes in the expectation of modeling errors. Based on experimental observations, the effectiveness of the system is demonstrated, while minimizing the impact of false alarms resulting from power supply imbalance, and it is shown that a robust and automatic electrical fault detection system has been produced.

Vibration Signal Analysis for Electrical Fault Detection of Induction Machine Using Neural Networks

Induction motors play an important role in the safe and efficient operation of industrial plants because they are considered inherently reliable due to their robust and relatively simple design. Therefore, there is an indispensable need to monitor their health and performance. And the detection and diagnosis of motor faults is the base to improve the efficiency of industrial plants. In this paper, a residual model-based fault detection method is proposed for steady state vibration signals of induction motors. The stationary signal had been extracted from the entire signal using data segmentation. The reference model in spectra is obtained statistically to represent the healthy condition. The ratio of RMS of residuals is used as a fault indicator for detecting motor faults. The effectiveness of the proposed approach in detecting a wide range of mechanical faults is demonstrated through the experiments applied on an 800 hp motor, and it is shown that a robust and reliable induction motor fault detection system has been produced.

Motor fault detection method for vibration signal using FFT residuals

Early detection and diagnosis of induction machine incipient faults are desirable for online condition monitoring, product quality assurance, and improved operational efficiency. However, conventional methods have to work with explicit motor models and cannot be used for vibration signal case because of their non-adaptation and the random nature of vibration signal. In this paper, a neural network method is developed for induction machine fault detection, using FFT. The neural network model is trained with vibration spectra and faults are detected from changes in the expectation of vibration spectra modeling error. The effectiveness and accuracy of the proposed approach in detecting a wide range of mechanical faults is demonstrated through staged motor faults, and it is shown that a robust and reliable induction machine fault detection system has been produced.

Hua Su

Papers

Induction Machine Condition Monitoring Using Neural Network Modeling

Vibration signal analysis for electrical fault detection of induction machine using neural networks

Vibration Signal Analysis for Electrical Fault Detection of Induction Machine Using Neural Networks

Motor fault detection method for vibration signal using FFT residuals

A neural network method for induction machine fault detection with vibration signal