Abstract
Although induction motors (IMs) are robust and reliable electrical machines, they can suffer different faults due to usual operating conditions such as abrupt changes in the mechanical load, voltage, and current power quality problems, as well as due to extended operating conditions. In the literature, different faults have been investigated; however, the broken rotor bar has become one of the most studied faults since the IM can operate with apparent normality but the consequences can be catastrophic if the fault is not detected in low-severity stages. In this work, a methodology based on convolutional neural networks (CNNs) for automatic detection of broken rotor bars by considering different severity levels is proposed. To exploit the capabilities of CNNs to carry out automatic image classification, the short-time Fourier transform-based time–frequency plane and the motor current signature analysis (MCSA) approach for current signals in the transient state are first used. In the experimentation, four IM conditions were considered: half-broken rotor bar, one broken rotor bar, two broken rotor bars, and a healthy rotor. The results demonstrate the effectiveness of the proposal, achieving 100% of accuracy in the diagnosis task for all the study cases.
Highlights
induction motors (IMs) is considered the most used electrical machine in industrial applications due to its features such as easy maintenance, great performance, low cost, and versatility [1]
Results in the convolutional neural networks (CNNs) complexity, a tradeoff between the information quantity that can be extracted from the analyzed image and the image size has to be established
It is worth noting that 25 pixels were selected as inputs for the 2D-CNN since they keep the information that is observed in the image size can be optimized by means of multi-objective optimization algorithms; the larger images but with a lower computation cost because the matrix size is reduced
Summary
IM is considered the most used electrical machine in industrial applications due to its features such as easy maintenance, great performance, low cost, and versatility [1]. To schedule maintenance times and avoid economic and human catastrophes, the development and application of diagnostic methods that offer more efficient and reliable results in terms of complexity and accuracy are still tasks of paramount importance, mainly considering BRB conditions at low severity, e.g., partially-broken rotor bars In this regard, many diagnosis methods based on diverse physical magnitudes such as current, vibration, ultrasound, temperature, and magnetic flux, among others, have been employed for identifying the BRB fault, being MCSA the most preferable magnitude because it allows measuring the physical characteristics of an IM without interrupting its normal operation [5,6,7,8]. Diverse works have focused on evaluating one or multiple BRBs, a consolidated fault (one or more bars completely segmented or cracked in two parts) [10,11,12]; few works have investigated a partially cracked bar, an initial condition of the BRB fault [9,13], because this condition alters slightly the monitored physical magnitudes, which increases the detection difficulty [14]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
