A Deep Learning Based on Sparse Auto-Encoder with MCSA for Broken Rotor Bar Fault Detection and Diagnosis

Convolutional Neural Network-Based Stator Current Data-Driven Incipient Stator Fault Diagnosis of Inverter-Fed Induction Motor

“…The basis for the operation of diagnostic systems based on neural networks (NNs), applied also for induction motor (IM) fault detection and classification, are analytical methods. Therefore, the input information for such systems is the result of extracting symptoms from chosen diagnostic signals using analytical methods, for example: FFT [24,25], WT [26], HHT [27], etc.…”

“…For the most part, they are associated with the analysis of mechanical damage to the IM [35,36] or mechanical system components [37,38]. A small number of works related to electrical damage of IM mainly concern rotor damages [24,27]. Most DNN-based systems use vibration measurements [39][40][41], less frequently stator currents [42][43][44] and voltages [45].…”

“…CNNs used in diagnostic applications are characterized by a much higher level of effectiveness compared to MLP and RBF networks, which are presented in [25,49]. In diagnostic processes, CNNs can act as a damage classification system [49], as well as provide information about the degree of damage [24]. In [51], the possibility of using CNNs as members of an autoencoder structure was presented.…”

“…An important aspect in the use of DNN in diagnostic processes is the appropriate adjustment of the measured signal to the structure and properties of the network. The input vector of DNNs can result from the signal analysis [24,25] and the directly provided diagnostic signal [35,46]. Due to the principle of operation of deep learning structures, in most cases the measured signal is converted into a 2D [25,48,52] or 3D [51] matrix.…”

Detecção de falhas com Stacked Autoencoders e técnicas de reconhecimento de padrões em poços de petróleo operados por gas lift

“…Diversas outras técnicas de Inteligência Computacional podem ser implementadas para detecção de falhas em processos industriais. Como o apresentado por Abdellatif et al (2018) que utiliza a saída de dois autoencoders como entradas de uma rede perceptron multicamadas (MLP) para detecção de barras quebradas em motores trifásicos de indução. No trabalho apresentado por Wen et al (2018) foram aplicados autoencoders em conjunto como uma rede neural MLP na detecção de falhas que causam desbalanceamento em turbinas de corrente marítima.…”