Applications of machine learning to machine fault diagnosis: A review and roadmap

Deep learning networks have been recently utilized for fault detection and diagnosis (FDD) due to its effectiveness in handling industrial process data, which are often with high nonlinearities and strong correlations. However, the valuable information in the raw data may be filtered with the layer-wise feature compression in traditional deep networks. This cannot benefit for the subsequent fine-tuning phase of fault classification. To alleviate this problem, an extended deep belief network (EDBN) is proposed to fully exploit useful information in the raw data, in which raw data is combined with the hidden features as inputs to each extended restricted Boltzmann machine (ERBM) during the pre-training phase. Then, a dynamic EDBN-based fault classifier is constructed to take the dynamic characteristics of process data into consideration. Finally, to test the performance of the proposed method, it is applied to the Tennessee Eastman (TE) process for fault classification. By comparing EDBN and DBN under different network structures, the results show that EDBN has better feature extraction and fault classification performance than traditional DBN.

A novel deep learning based fault diagnosis approach for chemical process with extended deep belief network.

A review of failure modes, condition monitoring and fault diagnosis methods for large-scale wind turbine bearings

Recent advances on SVM based fault diagnosis and process monitoring in complicated industrial processes

Signal based condition monitoring techniques for fault detection and diagnosis of induction motors: A state-of-the-art review

MLP neural network-based decision for power transformers fault diagnosis using an improved combination of Rogers and Doernenburg ratios DGA

In brushless excitation systems, the rotating diodes can experience open- or short-circuits. For a three-phase synchronous generator under no-load, we present theoretical development of effects of diode failures on machine output voltage. Thereby, we expect the spectral response faced with each fault condition, and we propose an original algorithm for state monitoring of rotating diodes. Moreover, given experimental observations of the spectral behavior of stray flux, we propose an alternative technique. Laboratory tests have proven the effectiveness of the proposed methods for detection of fault diodes, even when the generator has been fully loaded. However, their ability to distinguish between cases of diodes interrupted and short-circuited, has been limited to the no-load condition, and certain loads of specific natures.

/pdf/brushless-three-phase-synchronous-generator-under-rotating-4gc87oyzsn.pdf

Brushless Three-Phase Synchronous Generator Under Rotating Diode Failure Conditions

In this work we suggest an original fault signature based on an improved combination of Hilbert and Park transforms. Starting from this combination we can create two fault signatures: Hilbert modulus current space vector (HMCSV) and Hilbert phase current space vector (HPCSV). These two fault signatures are subsequently analysed using the classical fast Fourier transform (FFT). The effects of mechanical faults on the HMCSV and HPCSV spectrums are described, and the related frequencies are determined. The magnitudes of spectral components, relative to the studied faults (air-gap eccentricity and outer raceway ball bearing defect), are extracted in order to develop the input vector necessary for learning and testing the support vector machine with an aim of classifying automatically the various states of the induction motor.

Khmais Bacha

Papers

MLP neural network-based decision for power transformers fault diagnosis using an improved combination of Rogers and Doernenburg ratios DGA

Brushless Three-Phase Synchronous Generator Under Rotating Diode Failure Conditions

Support vector machine based decision for mechanical fault condition monitoring in induction motor using an advanced Hilbert-Park transform.

An improved combination of Hilbert and Park transforms for fault detection and identification in three-phase induction motors

EEMD-based notch filter for induction machine bearing faults detection