Application of Self-Organizing Neural Networks to Electrical Fault Classification in Induction Motors

Abstract

Electrical winding faults, namely stator short-circuits and rotor bar damage, in total constitute around 50% of all faults of induction motors (IMs) applied in variable speed drives (VSD). In particular, the short circuits of stator windings are recognized as one of the most difficult failures to detect because their detection makes sense only at the initial stage of the damage. Well-known symptoms of stator and rotor winding failures can be visible in the stator current spectra; however, the detection and classification of motor windings faults usually require the knowledge of human experts. Nowadays, artificial intelligence methods are also used in fault recognition. This paper presents the results of experimental research on the application of the stator current symptoms of the converter-fed induction motor drive to electrical fault detection and classification using Kohonen neural networks. The experimental tests of a diagnostic setup based on a virtual measurement and data pre-processing system, designed in LabView, are described. It has been shown that the developed neural detectors and classifiers based on self-organizing Kohonen maps, trained with the instantaneous symmetrical components of the stator current spectra (ISCA), enable automatic distinguishing between the stator and rotor winding faults for supplying various voltage frequencies and load torque values.

The Use of Artificial Neural Networks to Analyze Greenhouse Gas and Air Pollutant Emissions from the Mining and Quarrying Sector in the European Union

“…All neurons contain a specific number of weights (w), which is the same for each neuron, in accordance with the dimensions of the input vectors (x), i.e., the number of variables in the study. In other words, each neuron can be represented by a vector of dimensional (d) weights (w) = [w1, w2, ... wd], where d is the dimension of the input vectors (x) (Figure 4) [65].…”

Linear Method for Diagnosis of Inter-Turn Short Circuits in 3-Phase Induction Motors

“…The extended Park's vector approach (EPVA) was researched to improve the visualization of PVA so that the PVA could be analyzed in the form of an ellipse by applying the square root [33][34][35][36][37]. However, the research was conducted in a complex way by adding algorithms to algorithms (e.g., applying it to the fast Fourier transform, FFT).…”

“…On the one hand, from the view point of the harmonics, EPVA analysis of the second time harmonic in the frequency in order to detect the current supply unbalance is suggested, but data by MCSA is not easy to interpret because the harmonics of the hi-order generates as a lot of information is merged when the 3 phase is converted to 2 phase; there is the difficulty that additional work must be performed to identify harmonics [33][34][35][36][37]41].…”

“…However, the algorithm became too complex due to the complex circuit and needed to be simplified. Machine learning is a method later used for diagnosis with the recent development of artificial intelligence (AI) technology [18,23,30,37,[55][56][57][58][59]. AI-based methods are divided into feature extraction, fault identification, and fault severity evaluation.…”

Modified Rotor Flux Estimators for Stator-Fault-Tolerant Vector Controlled Induction Motor Drives

“…Many studies have focused on the incipient detection of ITSCs [10,11,14] for open-loop control schemes [15][16][17][18], as well as for closed-loop vector control structures of the IM. Field-oriented control (FOC) [19][20][21][22], direct torque control with switching table (DTC-ST) [23,24], and direct torque control with space-vector modulation (DTC-SVM) [21] structures have been considered.…”