A comparative study of deep learning efficiency in the classification of electrical faults of permanent magnet synchronous motor

Hiba Ziad; Ayad Q. Al-Dujaili; Amjad J. Humaidi

doi:10.1556/1848.2024.00885

International Review of Applied Sciences and Engineering

Volume/Issue:: Accepted Manuscript / Online First

A comparative study of deep learning efficiency in the classification of electrical faults of permanent magnet synchronous motor

Authors:

Hiba Ziad

Hiba Ziad Electrical Engineering Technical College, Middle Technical University, Baghdad 10022, Iraq

Search for other papers by Hiba Ziad in
Current site
Google Scholar
PubMed

bcc0048@mtu.edu.iq

Ayad Q. Al-Dujaili

Ayad Q. Al-Dujaili Electrical Engineering Technical College, Middle Technical University, Baghdad 10022, Iraq

Search for other papers by Ayad Q. Al-Dujaili in
Current site
Google Scholar
PubMed

ayad.qasim@mtu.edu.iq https://orcid.org/0000-0002-1126-3290

, and

Amjad J. Humaidi

Amjad J. Humaidi Control and Systems Engineering Department, University of Technology, Baghdad, Iraq

Search for other papers by Amjad J. Humaidi in
Current site
Google Scholar
PubMed

amjad.j.humaidi@uotechnology.edu.iq

View More View Less

Online Publication Date:: 04 Dec 2024

Publication Date:: 04 Dec 2024

Article Category:: Research Article

DOI:: https://doi.org/10.1556/1848.2024.00885

Keywords:: CWT; deep learning; fault classification; PMSM; scalogram

Open access

Download PDF

Abstract

Permanent Magnet Synchronous Motors (PMSMs) are widely used in modern industrial applications due to their high efficiency, reliability, and compact size. However, faults in PMSMs, such as stator winding failures, can lead to significant performance degradation and operational failures. Traditional fault detection methods often rely on signal processing and manual analysis, which may be time-consuming and lacking in accuracy. This study explores the application of deep learning techniques for automated fault detection in PMSMs. The deep learning models based on Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) are employed to classify electrical faults in the motor data, which includes the scalogram images of stator current signal allowing models to learn fault patterns. The performance of the used networks has been compared, in order to choose the reliable one for classification purposes and hence to be utilized for developing the prediction system. The experimental results show that the ResNet50 has better capability to classify the variation of data used where it could achieve 100% of accuracy, recall, precision, and F1 score as compared to other techniques.

Abstract

1 Introduction

One kind of motor that is well-known for its outstanding dynamic effectiveness and reliability is the permanent magnet synchronous motor (PMSM). This category of motor frequently operates under fluctuating conditions and experiences various transients such as load changes, frequent start/stop cycles, and acceleration/deceleration [1]. When PMSM is in use, performance may deteriorate or lead to failure, impacting the overall system's reliability and safety [2].

In most cases, the diagnosis of motor faults is divided into two main categories: those caused by electrical problems and those caused by mechanical problems [3, 4]. Some difficulties that can be seen in literature concerning manufacturing relate to problems of air-gap deformation, bearing failures, shaft misalignment, and mechanical imbalance [5–7]. Moreover, issues with the motor, electrical supply, and operating malfunctions were examined in the study [8–10].

Faults with the stator winding are frequently encountered and pose significant risks [11]. Mostly short circuits are a result of insulation damage. This damage can occur due to the age of the machine, during a time of significant flux weakening, because of a string of bad luck with weather and the state of the machine, because the winding is literally burning up from too much energy being put into it, or for any other number of reasons.

The first thing one might do to diagnose a problem with a turbine's stator winding is to look for short circuits between adjacent winding turns—a phenomenon called “inter-turn short circuit” (ITSC). ITSCs can rapidly escalate to more dangerous synchronous short circuits between phases or between a phase and ground. In the very early stages of the problem (when the ITSCs are still producing only very tiny changes in the amplitude of the phase currents), the motor may continue to operate safely. As the short circuiting proceeds, it will put the motor at risk. If these kinds of problems do develop, the only safe course of action—after making sure of what is happening—is to take the motor out of service [12]. It is critical to identify this fault quickly. The fault detection method for the motor has been around for more than a century, but it has been of particular interest recently due to the rapid advancement and extensive usage of PMSMs [13].

At present, electrical problems in PMSM are often detected by analyzing the stator's current in each phase. Several algorithms and diagnostic methods proposed to interpret the phase current for any fault indications [14]. Of these, the fast Fourier transform is one of the most commonly used frequency domain analysis techniques in motor current signature analysis (MCSA) applications [15]. Time-frequency domain methods consist of short-time Fourier transform (STFT) [16], and Continuous Wavelet Transform (CWT) [17]. Within the fault diagnosis domain, the CWT analysis efficacy has primarily been explored with induction motors [18]. Nevertheless, notwithstanding its numerous advantages, this facet has not been previously investigated concerning its capacity to identify electrical malfunctions in PMSMs during their nascent phase. This study aims to assess the feasibility of using signal preprocessing to differentiate between the indications of ITSCs in the PMSM stator winding.

The signals processed can be utilized to train AI fault classifier models. Lately, this method has become widely adopted for automating the electric motor condition monitoring process. Artificial neural networks (ANNs) are frequently utilized in fault diagnosis. Neural networks with shallow structures, like multilayer perceptron (MLP) [19], and self-organizing Kohonen maps (SOMs) [20] have been quite popular. Nevertheless, similar applications in recent years have widely embraced trained convolutional neural networks (CNNs). These networks use transfer learning approaches, as well as deep (and therefore large) neural networks like Long Short-Term Memory (LSTM) [21–25 ]. It is important to highlight that by utilizing a suitable diagnostic signal and pre-processing technique, a basic and user-friendly machine learning (ML) algorithm like Support Vector Machine (SVM) could be effective [26].

The entire experiment mainly utilizes the frequency-domain characteristic of these features to detect the fault for a PMSM at different operation conditions. The fault frequency fluctuates due to the variation of fundamental frequency of stator current, which also change depends on motor operating condition. More importantly, as the primary focus of this paper is to find an efficient TL model that can classify the electrical faults accurately, bulk failures in terms of their frequency, and stator winding short-turn faults with inter-turn short circuit fault being the most prevalent type. Broadly speaking there are four basic kinds of electrical malfunction; ground faults, short circuits, connections errors, and open circuits. The Inter-Turn Short Fault is the most common of these, and it will be the only one studied in this paper. For each simulation of this cluster, the current signal measurement will be treated with a CWT operator. In other words, we should end up with an image that would keep meaningful elements of our original signal. After the defective and healthy feature libraries have been created, a total of nine different kinds of Deep Neural Networks will be developed for division in Fig. 1.

Fig. 1.

Procedure for identifying errors

Citation: International Review of Applied Sciences and Engineering 2025; 10.1556/1848.2024.00885

The study's primary novelty can be outlined as follows:

To the best of the author's knowledge, there is no such published work which discuses using complex generalized CWT analysis of scalogram obtained by processing diagnostic signals-based information for detecting electrical faults from stator winding using PMSM. This paper aims to remedy that shortcoming.
The second objective is to introduce a method for automatically classifying PMSM stator winding faults using specific Deep Learning Networks. Conducting a comparison of nine distinct designs - ResNet18, ResNet50, ResNet101, Vgg16, Vgg19, GoogleNet, ShuffleNet, EfficientNetb0, and MobileNet-V2, involves analyzing the impact of model parameters on the performance of the classifiers. Analyzing such data is frequently disregarded, yet it can prove highly beneficial when developing a diagnostic system.

Here is the remaining structure of this paper: The first section includes an overview of the subject and a literature review; the second section details the technique that will be used; and the third section contains the findings. Section 4 wraps up the report by outlining potential areas for further investigation.

2 Materials and methods

Most of the time, the most common electric faults in windings have to do with poor grounding or with connections that are not made. These can lead to short circuits. By the time, a phase winding that should be connected to a load may be found wanting. The open circuit itself might be the electrical fault [27]. Stator faults are responsible for 38% of all motor defects [28]. The PMSM systems commonly suffer from a problem, which is the short circuits that occur between turns of the stator winding [29]. When an electric motor's working time increases, defects can develop in the stator wire insulation system. A short circuit is one of the most serious defects that can occur. A short circuit between the stator turns can stop the motor and possibly cause a fire if the problem is not discovered and fixed [30].

A short circuit fault is challenging to manage due to the generation of a substantial circulating current and excessive heat along the shorted path [31]. Consequently, failure to promptly detect and eradicate this phenomenon can lead to further detrimental effects on wire insulation and rapid propagation into additional stator windings. Such propagation can result in demagnetization, phase-to-phase faults, and even ground faults [32].

An open circuited phase may result from a malfunctioning internal stator winding [33], although this is frequently the result of drive system issues. The motor can have really strong mechanical vibrations and changes in electromagnetic torque. This can lead to additional problems if the motor is started up and allowed to run while the coils are shorted. So the first coil winding to reach the arms of the rotor would not be working, and the rest would be checked to follow [34]. The descriptions of a PMSM, such as vibration, current, voltage, torque, and magnetic flux, might change due to flaws. Nonetheless, the stator phase current signal conveys more advantageous information. Figure 2 illustrates the PMSM simulation model, the approach extraction and classification processes, and the SimPower system toolbox in Matlab/Simulink, which were utilized to record phase current data.

Fig. 2.

Flowchart of proposed classification method

Citation: International Review of Applied Sciences and Engineering 2025; 10.1556/1848.2024.00885

Data creation, feature extraction, and the signal-image conversation technique are some of the ideas introduced in the following sections that form the basis of the suggested categorization approach.

2.1 Data collection

Unlike the above studies, these results have been acquired by conducting experiments on the PMSM at a normal operating condition. Those details for the motor are registered in Table 1. The missing of any step-like amplitude change in the present phase signal, and steady running speed of motor indicates that motor is working properly. A leg inverter connected to 500 V DC was used to feed the three-phase PMSM which was also adopted as experimental model. However, the PMSM used in this work can be expressed as: The stator has a resistance of 2.8 Ω, leakage inductance of 0.42 H and flux linkage of 0.175 Wb with the rated speed of 3,000 RPM. There are 3 poles, with a moment of inertia (J) = 0.03 kg m⁻².

Table 1.

Faults created on inverter side

Phase	IGBT1	IGBT2	IGBT3	IGBT4	IGBT5	IGBT6
Phase a	O	I	I	I	I	I
Phase a	I	I	I	O	I	I
Phase b	I	I	O	I	I	I
Phase b	I	I	I	I	I	I
Phase c	I	I	I	I	O	I
Phase c	I	O	I	I	I	I
Phase a, c	O	O	I	I	I	I
Phase a, b	O	I	I	I	I	O
Phase a, b	I	I	O	O	I	I
Phase b, c	I	O	O	I	I	I
Phase a, c	I	I	I	O	O	I
Phase b, c	I	I	I	I	I	O
Phase a, b	O	I	O	I	I	I
Phase b, c	I	I	O	I	O	I
Phase a, c	O	I	I	I	O	I
Phase a, c	I	I	I	O	I	O
Phase b, c	I	O	I	I	I	O
Phase a, b	I	O	I	O	I	I
Phase a, b, c	O	I	O	I	O	I
Phase a, b, c	I	O	I	O	I	O
Phase a, b, c	O	I	I	I	O	O
Phase a, b, c	I	O	O	O	I	I
Phase a, b, c	O	O	I	O	I	O
Phase a, b, c	I	O	O	O	I	O
Phase a, b, c	I	O	I	O	O	O
Phase a, b, c	O	I	O	O	O	I
Phase a, b, c	O	I	O	I	O	O
Phase a, b, c	O	O	O	I	O	I

Then a lookup table is shown built to account for the openings and shorts circuits faults on the drive side (which are simulated after obtaining different situations of conditions in PMSM – see Table 1) so that fault signals can be considered. In the motor fault model, expediently after 0.2 s of runtime there are coil-to-ground and coil-coil defects as shown from now on the lowest possible value of the fault resistance was applied before abrupt changes in the phase current signal took place. This is reflected by a nine-column vector with ones only set at five positions [35].

With 476 labels total, one for every category in the dataset. In order to train deep learning models more efficiently and with less overfitting, preparation is necessary regardless of whether the dataset has already been retrieved or not. Unfortunately, the method fails when it comes to recognizing the minority group because of bias within the ruling elite. A data balancing is performed to ensure this does not occur. Similarly, data augmentation techniques were used to supplement the quantity with a modified version of the original dataset. On top of that, we used five distinct data augmentation methods—reflection, vertical cropping, horizontal cropping, and random rotation—to the training dataset alone. These methods helped improve the model's accuracy, along with epoch-by-epoch random dataset selection.

2.2 Scalogram image acquisition

Wavelet transform has been a powerful tool for improving the signal-to-noise ratio and it is best applied on 2D photos of time-domain signals, which is suitable to nonstationary and nonlinear data. The basic functions of the wavelet transform are among an infinite number that are possible, as opposed to those of the Fourier transform. Wavelet analysis used for signal data gives a more accurate result as compared to other signal analysing techniques. It examines time and frequency-localized features simultaneously with high precision [36, 37].

Continuous or discrete, wavelet analysis is employed according to the purpose for which the data is to be used [38]. It is recommended to use CWT for in-depth time-frequency analysis. The scale parameter is discretized differently in continuous and discontinuous wavelet transformations. CWT often employs a base less than 2 exponential scale, in contrast to the constant usage of a base equal to 2 exponential scale in the discrete wavelet transform. Consequently, CWT is superior to the discrete wavelet transform in terms of scale discretization.

When dealing with nonstationary signals, CWT works better for pinpointing transients. In terms of detecting the frequency content of a signal, CWT outperforms the discrete wavelet approach. Time resolution in each frequency band is same between the original data and CWT [39]. In addition, CWT is invariant with respect to displacement, whereas DWT is dependent on it. When it comes to nonstationary signal analysis, CWT is the time-frequency transform that works best. The continuous wavelet transform is defined by the equation given here. This equation expresses the transform as a kind of correlation between the signal being transformed—which is thought of as a function of continuous time, x(t)—and a kernel (or wavelet function) that is also defined as a function of time [40].

C w (a, b) = \frac{1}{\sqrt{a}} \int_{- \infty}^{+ \infty} x (t) ψ * \frac{(t - b)}{a} d t

(1)

From the above equation, we can infer that ψ refers to our continuous mother wavelet. It is translated using a scaling factor (b) and ‘a’ represent Scaling factor i.e. resizes or shrinks it, Cw (a, b) subsequently becomes a function over two parameters: By observing that here values of j and k are also continuous means number of Wavelets will be infinite. The complex conjugate of the mother that we are analysing is φ(t) and, consequently its dual-scale function: ψ*(t). Factor 1/√a is an energy-normalized coefficient. To further qualify as a legitimate ground wavelet, the wavelet function must satisfy the subsequent mathematical requirements:

There must be a limited quantity of energy in the wavelet.

E = \int_{- \infty}^{+ \infty} {| ψ (t) |}^{2} d t < \infty

(2)

C_{ψ} = \int_{- \infty}^{+ \infty} \frac{{| φ (w) |}^{2}}{w} d w < \infty

(3)

where,

φ (w) = \int_{- \infty}^{+ \infty} ψ (t) e^{- i ω t} d t

(4)

The function φ (t) is transformed using the Fourier transform, and the circular frequency is represented by w = 2πf. [41].

Scalograms are spectra of amplitude and time scale that result from the application of CWTs to signals. Scalograms show the absolute value of the CWT signal as a function of time and frequency. The end product of CWT is a two-dimensional function ${E (a, b) = | Cw (a, b) |}^{2}$ , which is known as a scaleogram signal energy distribution for parameters a and b [42]. In contrast to spectrograms, which involve windowing the input signal through the shifting of a window of constant duration in time and frequency, the application of a scalogram to analyse real world signals containing characteristics that manifest at distinct scales is more practical.

The starting point in CWT is often set to a positive integer, such as, $2^{1 / v}$ , where v is a parameter called the ‘voices per octave.’ The number of wavelet filters per octave is often referred to by this name. In a multiplicative meaning, “voices” are “the scale levels between subsequent octaves.” [43]. The quantity of octaves present in an audio or music signal establishes the range of frequencies that are intended for analysis. The abundance of samples (scale) within a given range is ascertained by the voice count per octave [44]. Distinct scales result from increasing the magnitude of this base scale by positive integer powers. The number of voices per octave (v) derives its nomenclature from the fact that v intermediate scales are required for a one-octave duplication of the scale. Values of v that occur frequently include 32, 10, 12, 14, and 16. As v grows, the degree of accuracy in discretizing the scale parameter improves. However, this causes an expansion in computational time due to the necessity of calculating the CWT for every scale [45].

The time-frequency image, which is a two-dimensional representation of the energy intensity of the impact signals, possesses high instantaneous energy properties. This facilitates the application of the Morlet wavelet as a wavelet function.

2.3 Feature extraction and image classification

The classification of stator current scalogram images was accomplished utilizing the transfer learning-based CNN algorithm implemented in this study. The major purpose of this investigation is to determine which pre-trained CNN model is most likely to be successful. In practice, CNN algorithms are implemented on large datasets as opposed to tiny ones. Transfer learning principles may prove advantageous in situations where only a limited dataset is accessible, and a model that has been trained on larger datasets can produce satisfactory results when applied to the smaller data. Transfer learning has recently been implemented successfully in numerous industries, including medical image classification, manufacturing, and baggage screening [46].

In order to classify the stator current scalogram pictures, some pre-trained deep convolutional neural network models are utilized: ResNet18, ResNet50, ResNet101, Vgg16, V19, GoogleNet, Efficientnetb0, MobileNetv2, and ShuffleNet are all versions of the ResNet network. In Tables 2 and 3, the number of layers and magnitude of input for each structure and training parameter are detailed, respectively.

Table 2.

The total number of layers and the size of the input layer for each network

Network	Input layer Size	No. of Layer
ResNet18	224*224	72
ResNet50	224*224	177
ResNet101	224*224	347
Vgg16	224*224	41
Vgg19	224*224	47
GoogleNet	224*224	144
Efficientnetb0	224*224	237
MobileNetv2	224*224	28
ShuffleNet	224*224	48

Table 3.

Training parameters for each network

Software	Network	Learning rate	Mini batch size	Training: Test ratio
Matlab	ResNet18	0.1, 0.01, 0.001, 0.0001	32, 16, 8	60:40, 70:30, 80:20
	ResNet50	0.1, 0.01, 0.001, 0.0001	32, 16, 8	60:40, 70:30, 80:20
	ResNet101	0.1, 0.01, 0.001, 0.0001	32, 16, 8	60:40, 70:30, 80:20
	Vgg16	0.1, 0.01, 0.001, 0.0001	32, 16, 8	60:40, 70:30, 80:20
	Vgg19	0.1, 0.01, 0.001, 0.0001	32, 16, 8	60:40, 70:30, 80:20
	GoogleNet	0.1, 0.01, 0.001, 0.0001	32, 16, 8	60:40, 70:30, 80:20
	Efficientnetb0	0.1, 0.01, 0.001, 0.0001	32, 16, 8	60:40, 70:30, 80:20
	MobileNetv2	0.1, 0.01, 0.001, 0.0001	32, 16, 8	60:40, 70:30, 80:20
	ShuffleNet	0.1, 0.01, 0.001, 0.0001	32, 16, 8	60:40, 70:30, 80:20

The RGB image is pre-processed to be a 224 × 224 picture classifying it into either Healthy/Faulty using Neural Network. Now, in order to get rid of the use of pre-trained models having an ImageNet architecture we have no other option but that our data allows us to build those dense and dropout levels from scratch in the classification layers: Nevertheless, it was that, but our training dataset is fed to the convolution layer of ResNet model and after getting those output vectors of final layer. The output vectors of the final layer are shown to the Softmax classifier (classification layer), using that as training data. The experiment is done using a Matlab environment with only single CPU of use. During the beginning of each training cycle a random subset from the original dataset was taken as the trained dataset. During the first stage, three parts were considered; that is 60%, 70%, and 80% of the subsets. 10 epochs were used by default as shown below: here the learning rates were 0.1, 0.01, 0.01, 0.001 and 0.0001, respectively, which was done using adaptive moment estimation (Adam). Thus, we started with 32/16/8 batches for training.

2.4 Performance evaluation

All of the identified classes' real true positives, actual true negatives, false positives, and false negatives are combined in a single table employing a confusion matrix. This is done so that it gives an unbiased assessment of the overall performance of a classification model when tested on an independent test set. To construct a confusion matrix, you first compare the class labels that we saw in the test dataset to the ones our model was predicting. Where the rows represent the actual class labels and columns are exposed into the anticipated ones. Despite the above, we evaluated this experiment's case with our suggested approach by determining how different the actual categories were from what was classified by the classifier. To compute mathematically, this operation is called accuracy; or better written as:

A C C = \frac{T P + T N}{T P + T N + F N + F P}

(5)

In the case of true positive (TP), our model accurately predicted those cases which have actually a positive 306 transaction. And a model that has correctly been able to identify these negative scenarios were the instances which are actually negative and also detected by the model. This is called true negative (TN). Also, false positives (FP) occur when the model mistakenly classifies a case as positive, while false negatives (FN) represent cases where the model incorrectly predicts a positive instance as negative.

Further, ‘Recall’ measures how well a model can identify all the positive cases. It is given by the ratio of total number of true positive predictions produced from a model to total number of actual positive predictions:

R e c a l l = \frac{T P}{T P + F N}

(6)

The definition of precision is to what degree the model can detect true positive cases. It is computed as the fraction of TP predictions to all predicted positive instances:

P r e c i s i o n = \frac{T P}{T P + F P}

(7)

Finally, we have the F1 score, which is computed as follows and signifies the system's ability to balance accuracy and recall:

F 1 S c o r e = \frac{2 * P r e c i s i o n * R e c a l l}{P r e c i s i o n + R e c a l l}

(8)

3 Results

In this research, nine of deep learning models were utilized in this comparative analysis to classify electrical defects of the stator current in PMSM. This research aimed to study the performance and make a competition for these models by using crucial characteristics that recognize the efficacy of the classification system. The accuracy, recall, precision, and F score comprise these parameters. ResNet18, ResNet50, ResNet101, Vgg16, Vgg19, GoogleNet, EfficientNetb0, MobileNetv2, and ShuffleNet comprised the nine models. Table 4 and Fig. 3 below show the best-achieved results and performance evaluation of all nine DL models trained in this research. However, it is very important to compare the confusion matrix of best training step for each network Fig. 4. This will allow to figure out networks ability to classify the features of faults.

Table 4.

Best achieved result for every transfer-learning model

Transfer-learning model	Evaluation metrics
Transfer-learning model	Accuracy	Recall	Precision	F-Score
ResNet18	99.3	100	98.5	99.2
ResNet50	100	100	100	100
ResNet101	98.6	98.6	98.6	98.6
Vgg16	94.6	100	89.3	94.3
Vgg19	94.6	92	97.8	94.8
GoogleNet	98.9	100	97.8	98.8
EfficientNetb0	99.4	100	98.5	99.4
MobileNetv2	98.9	100	97.8	98.8
ShuffleNet	99.3	100	98.5	99.2

Fig. 3.

Performance evaluation for every trained model

Citation: International Review of Applied Sciences and Engineering 2025; 10.1556/1848.2024.00885

Fig. 4.

Confusion matrices of ResNet18, ResNet50, ResNet101, Vgg16, Vgg19, GoogleNet, ShuffleNet, EfficientNetb0, and MobileNetv2 respectively

Citation: International Review of Applied Sciences and Engineering 2025; 10.1556/1848.2024.00885

4 Discussion

The performance of all nine TL models was good with our limited data, but ResNet50 shows a significant and unique accuracy percentage, where it achieves 100% when trained on 80% of the data, mini-batch 8, and a learning rate of 0.0001. Figure 4 represents the confusion matrices of nine trained models.

However, the rest of the TL model achieved a good accuracy percentage when compared with the best achieved one which is ResNet50 in our case, but one can notice the differences with the other evaluation metrics, for example, F-score which is the most important one that indicates the potentiality of the model in balancing precision, and recall harmonic mean. ResNet50 is ranked first on our list with 100%, followed by EfficientNetb0, which achieved 99.4%, then both ResNet18 and ShuffleNet, which obtained 99.2%, GoogleNet and MobileNetv2 with 98.8%, ResNet101 with 98.6%, Vgg19 with 94.8%, and finally Vgg16, which achieved 93.2%. Nonetheless, the results of ResNet50 shown in this paper give new perspectives into the feasibility of selecting such network for online electrical fault diagnosis. This makes the remote monitoring of motor performance easier and predicts what type of fault is going to happen so that necessary action may be taken before it happens. Still, even in the worst circumstances in whatever line of plant machinery you use to get your work done, it remains very possible to lower any chance that the motor would be impaired (stopping a certain manufacturing line), or other parts wrecked instead.

As shown in the results section, ResNet 50 achieved 100% in all metrics, this can be referred to the fact of its 50 residual blocks, which has proven extraordinary capability in overcoming vanishing gradient issues, enabling the effective training of deeper neural networks. However, Residual Neural Networks changed the training procedure of Deep Neural Networks in classification manner. The basic strategy depends on skip connections and residual blocks, however, these blocks are not only changed the way of training these networks, but has also accelerated the construction of more complex and efficient models.

Furthermore, more layers can make the model more complex and add more capacity, and this does not make the model perform better, but it can make it struggle to converge. In case it can escape from converge, the model can also go too far and over-fit, which means the model can perform very well with training set, but will confused with testing set.

The classification of faults can be employed or merged with different control strategies to establish fault-tolerant control schemes [47–65]. Moreover, other extension of this work is to implement in real-time environment using either FPGA (Field Programming Gate Array) format, Raspberry-pi microcomputer, or NI board using LabView software. [66–73].

5 Conclusion

This study established the basis to develop an early fault-detection system. In this research, nine TL models have been used to classify the electric faults of PMSM based on status of stator currents. Different deep learning techniques are developed for automated fault detection in PMSMs. The proposed deep learning models, which are based on Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) are employed to classify electrical faults in the motor data which includes the scalogram images of stator current signal allowing models to learn fault patterns. The performance of the used networks has been compared to choose the reliable one for classification purposes. The results showed that the ResNet50 has better capability to classify the variation of data used where it could achieve 100% of accuracy, recall, precision, and F1 score, as compared to other techniques. There is a potential future development of this study to add more kinds of faults to enrich the dataset with sufficient diversity, so that more features can be extracted to train the model. This work could assist the researchers in the development of fault-detection system of motor based on the status of stator current.

References

[1]↑
J. A. Antonino-Daviu, A. Quijano-López, M. Rubbiolo, and V. Climente-Alarcon, “Advanced analysis of motor currents for the diagnosis of the rotor condition in electric motors operating in mining facilities,” IEEE Trans. Industry Appl., vol. 54, no. 4, pp. 3934–3942, July–Aug. 2018. https://doi.org/10.1109/TIA.2018.2818671.
- Search Google Scholar
- Export Citation
[2]↑
Z. Ullah and J. Hur, “A comprehensive review of winding short circuit fault and irreversible demagnetization fault detection in PM Type Machines,” Energies, vol. 11, no. 12, p. 3309, Nov. 2018. https://doi.org/10.3390/en11123309.
- Search Google Scholar
- Export Citation
[3]↑
P. Gangsar and R. Tiwari, “Signal based condition monitoring techniques for fault detection and diagnosis of induction motors: a state-of-the-art review,” Mech. Syst. Signal Process., vol. 144, Oct. 2020, Art no. 106908. https://doi.org/10.1016/j.ymssp.2020.106908.
- Search Google Scholar
- Export Citation
[4]↑
X. Liang, M. Z. Ali, and H. Zhang, “Induction motors fault diagnosis using finite element method: a review,” IEEE Trans. Industry Appl., vol. 56, no. 2, pp. 1205–1217, Mar. 2020. https://doi.org/10.1109/tia.2019.2958908.
- Search Google Scholar
- Export Citation
[5]↑
N. Yassa, M. Rachek, and H. Houassine, “Motor current signature analysis for the air gap eccentricity detection in the squirrel cage induction machines,” Energy Proced., vol. 162, pp. 251–262, Apr. 2019. https://doi.org/10.1016/j.egypro.2019.04.027.
- Search Google Scholar
- Export Citation
[6]
C. Abdelkrim, M. S. Meridjet, N. Boutasseta, and L. Boulanouar, “Detection and classification of bearing faults in industrial geared motors using temporal features and adaptive neuro-fuzzy inference system,” Heliyon, vol. 5, no. 8, Aug. 2019. https://doi.org/10.1016/j.heliyon.2019.e02046.
- Search Google Scholar
- Export Citation
[7]
C. Li, J. Xiong, X. Zhu, Q. Zhang, and S. Wang, “Fault diagnosis method based on encoding time series and Convolutional Neural Network,” IEEE Access, vol. 8, pp. 165232–165246, 2020. https://doi.org/10.1109/access.2020.3021007.
- Search Google Scholar
- Export Citation
[8]↑
S. M. Zaman and X. Liang, “An effective induction motor fault diagnosis approach using graph-based semi-supervised learning,” IEEE Access, vol. 9, pp. 7471–7482, 2021. https://doi.org/10.1109/access.2021.3049193.
- Search Google Scholar
- Export Citation
[9]↑
A. Jafari, J. Faiz, and M. A. Jarrahi, “A simple and efficient current-based method for Interturn fault detection in BLDC Motors,” IEEE Trans. Ind. Inform., vol. 17, no. 4, pp. 2707–2715, Apr. 2021. https://doi.org/10.1109/tii.2020.3009867.
- Search Google Scholar
- Export Citation
[10]
T. A. Shifat and J.-W. Hur, “Ann assisted multi sensor information fusion for BLDC motor fault diagnosis,” IEEE Access, vol. 9, pp. 9429–9441, 2021. https://doi.org/10.1109/access.2021.3050243.
- Search Google Scholar
- Export Citation
[11]↑
J. He, C. Somogyi, A. Strandt, and N. A. Demerdash, “Diagnosis of stator winding short-circuit faults in an interior permanent magnet synchronous machine,” in 2014 IEEE Energy Conversion Congress and Exposition (ECCE), Sep. 2014. https://doi.org/10.1109/ecce.2014.6953825.
- Search Google Scholar
- Export Citation
[12]↑
M. Drif and A. J. Cardoso, “Stator fault diagnostics in squirrel cage three-phase induction motor drives using the instantaneous active and reactive power signature analyses,” IEEE Trans. Ind. Inform., vol. 10, no. 2, pp. 1348–1360, May 2014. https://doi.org/10.1109/tii.2014.2307013.
- Search Google Scholar
- Export Citation
[13]↑
Z. Wang, J. Yang, H. Ye, and W. Zhou, “A review of permanent magnet synchronous motor fault diagnosis,” in 2014 IEEE Conference and Expo Transportation Electrification Asia-Pacific (ITEC Asia-Pacific), Beijing, China, 2014, pp. 1–5. https://doi.org/10.1109/ITEC-AP.2014.6940870.
- Search Google Scholar
- Export Citation
[14]↑
E. A. Bhuiyan, M. M. Azad Akhand, S. K. Das, F. Ali, Z. Tasneem, R. Islam, D. K. Saha, F. R. Badal, H. Ahamed, and S. I. Moyeen, “A survey on fault diagnosis and fault tolerant methodologies for permanent magnet synchronous machines,” Int. J. Automation Comput., vol. 17, no. 6, pp. 763–787, Nov. 2020. https://doi.org/10.1007/s11633-020-1250-3.
- Search Google Scholar
- Export Citation
[15]↑
R. Z. Haddad and E. G. Strangas, “On the accuracy of fault detection and separation in permanent magnet synchronous machines using MCSA/MVSA and LDA,” IEEE Trans. Energy Convers., vol. 31, no. 3, pp. 924–934, Sept. 2016. https://doi.org/10.1109/TEC.2016.2558183.
- Search Google Scholar
- Export Citation
[16]↑
P. Pietrzak and M. Wolkiewicz, “Stator phase current STFT analysis for the PMSM stator winding fault diagnosis,” in 2022 International Symposium on Power Electronics, Electrical Drives, Automation and Motion (SPEEDAM), Jun. 2022. https://doi.org/10.1109/speedam53979.2022.9841990.
- Search Google Scholar
- Export Citation
[17]↑
M. Al Duhayyim, G. Mohamed, S. Alzahrani, R. Alabdan, A. S. A. Aziz, A. S. Zamani, I. Yaseen, and M. I. Alsaid, “Sandpiper optimization with a deep learning enabled fault diagnosis model for complex industrial systems,” Electronics, vol. 11, no. 24, p. 4190, Dec. 2022. https://doi.org/10.3390/electronics11244190.
- Search Google Scholar
- Export Citation
[18]↑
S. M. Zaman, H. U. Marma, and X. Liang, “Broken rotor bar fault diagnosis for induction motors using power spectral density and complex continuous wavelet transform methods,” in 2019 IEEE Canadian Conference of Electrical and Computer Engineering (CCECE), May 2019. https://doi.org/10.1109/ccece.2019.8861517.
- Search Google Scholar
- Export Citation
[19]↑
M. Skowron, M. Krzysztofiak, and T. Orlowska-Kowalska, “Effectiveness of neural fault detectors of permanent magnet synchronous motor trained with symptoms from field-circuit modeling,” IEEE Access, vol. 10, pp. 104598–104611, 2022. https://doi.org/10.1109/access.2022.3211087.
- Search Google Scholar
- Export Citation
[20]↑
M. Skowron, T. Orlowska-Kowalska, and C. T. Kowalski, “Diagnosis of stator winding and permanent magnet faults of PMSM drive using shallow neural networks,” Electronics, vol. 12, no. 5, p. 1068, Feb. 2023. https://doi.org/10.3390/electronics12051068.
- Search Google Scholar
- Export Citation
[21]↑
Q. Song, M. Wang, W. Lai, and S. Zhao, “On bayesian optimization-based residual CNN for estimation of Inter-Turn Short Circuit Fault in PMSM,” IEEE Trans. Power Electronics, vol. 38, no. 2, pp. 2456–2468, Feb. 2023. https://doi.org/10.1109/tpel.2022.3207181.
- Search Google Scholar
- Export Citation
[22]↑
M. Skowron, T. Orlowska-Kowalska, and C. T. Kowalski, “Detection of permanent magnet damage of PMSM drive based on direct analysis of the stator phase currents using convolutional neural network,” IEEE Trans. Ind. Electronics, vol. 69, no. 12, pp. 13665–13675, Dec. 2022. https://doi.org/10.1109/tie.2022.3146557.
- Search Google Scholar
- Export Citation
[23]↑
C. Parkash, Y. Zhou, A. Kumar, G. Vashishtha, H. Tang, and J. Xiang, “A novel entropy-based sparsity measure for prognosis of bearing defects and development of a sparsogram to select sensitive filtering band of an axial piston pump,” SSRN Electron. J., 2022. https://doi.org/10.2139/ssrn.4182162.
- Search Google Scholar
- Export Citation
[24]↑
M. Skowron and C. T. Kowalski, “Permanent magnet synchronous motor fault detection system based on transfer learning method,” in IECON 2022 – 48th Annual Conference of the IEEE Industrial Electronics Society, Oct. 2022. https://doi.org/10.1109/iecon49645.2022.9968867.
- Search Google Scholar
- Export Citation
[25]
H. Kaplan, K. Tehrani, and M. Jamshidi, “A fault diagnosis design based on deep learning approach for electric vehicle applications,” Energies, vol. 14, no. 20, p. 6599, Oct. 2021. https://doi.org/10.3390/en14206599.
- Search Google Scholar
- Export Citation
[26]↑
Y.-P. Zhao, G. Huang, Q.-K. Hu, and B. Li, “An improved weighted one class support vector machine for Turboshaft Engine Fault Detection,” Eng. Appl. Artif. Intelligence, vol. 94, Sep. 2020, Art no. 103796. https://doi.org/10.1016/j.engappai.2020.103796.
- Search Google Scholar
- Export Citation
[27]↑
S. Nandi, H. A. Toliyat, and X. Li, “Condition monitoring and fault diagnosis of Electrical Motors—a review,” IEEE Trans. Energy Convers., vol. 20, no. 4, pp. 719–729, Dec. 2005. https://doi.org/10.1109/tec.2005.847955.
- Search Google Scholar
- Export Citation
[28]↑
S. S. Moosavi, Q. Esmaili, A. Djerdir, and Y. A. Amirat, “Inter-turn fault detection in stator winding of PMSM using wavelet transform,” 2017 IEEE Vehicle Power Propulsion Conf. (VPPC), vol. 29, pp. 1–5, Dec. 2017. https://doi.org/10.1109/vppc.2017.8330891.
- Search Google Scholar
- Export Citation
[29]↑
M. A. Mazzoletti, G. R. Bossio, C. H. De Angelo, and D. R. Espinoza-Trejo, “A model-based strategy for Interturn short-circuit fault diagnosis in PMSM,” IEEE Trans. Ind. Electronics, vol. 64, no. 9, pp. 7218–7228, Sep. 2017. https://doi.org/10.1109/tie.2017.2688973.
- Search Google Scholar
- Export Citation
[30]↑
F. Cira, M. Arkan, B. Gumus, and T. Goktas, “Analysis of stator inter-turn short-circuit fault signatures for inverter-fed permanent magnet synchronous motors,” in IECON 2016 – 42nd Annual Conference of the IEEE Industrial Electronics Society, vol. 55, Oct. 2016, pp. 1453–1457. https://doi.org/10.1109/iecon.2016.7793717.
- Search Google Scholar
- Export Citation
[31]↑
Y. Liang, “Diagnosis of inter-turn short-circuit stator winding fault in PMSM based on stator current and noise,” in 2014 IEEE International Conference on Industrial Technology (ICIT), Feb. 2014. https://doi.org/10.1109/icit.2014.6894927.
- Search Google Scholar
- Export Citation
[32]↑
K.-H. Kim, “Simple online fault detecting scheme for short-circuited turn in a PMSM through current harmonic monitoring,” IEEE Trans. Ind. Electronics, vol. 58, no. 6, pp. 2565–2568, Jun. 2011. https://doi.org/10.1109/tie.2010.2060463.
- Search Google Scholar
- Export Citation
[33]↑
K.-C. Kim, S.-B. Lim, D.-H. Koo, and J. Lee, “The shape design of permanent magnet for permanent magnet synchronous motor considering partial demagnetization,” IEEE Trans. Magnetics, vol. 42, no. 10, pp. 3485–3487, Oct. 2006. https://doi.org/10.1109/tmag.2006.879077.
- Search Google Scholar
- Export Citation
[34]↑
B. A. Welchko, T. M. Jahns, and S. Hiti, “IPM synchronous machine drive response to a single-phase open circuit fault,” IEEE Trans. Power Electronics, vol. 17, no. 5, pp. 764–771, Sep. 2002. https://doi.org/10.1109/tpel.2002.802180.
- Search Google Scholar
- Export Citation
[35]↑
H. Z. Faraj, A. Q. AL-Dujaili, and A. J. Humaidi, “The classification method of electrical faults in permanent magnet synchronous motor based on deep learning,” in 2023 IEEE 11th Conference on Systems, Process & Control (ICSPC), Malacca, Malaysia, 2023, pp. 326–331. https://doi.org/10.1109/ICSPC59664.2023.10420154.
- Search Google Scholar
- Export Citation
[36]↑
J. Hang, J. Zhang, and M. Cheng, “Detection and discrimination of open phase fault in permanent magnet synchronous motor drive system,” IEEE Trans. Power Electronics, pp. 1–1, 2015. https://doi.org/10.1109/tpel.2015.2479399.
- Search Google Scholar
- Export Citation
[37]↑
N. Bhatnagar, “Introduction to wavelet transforms,” Feb. 2020. https://doi.org/10.1201/9781003006626.
- Search Google Scholar
- Export Citation
[38]↑
L. Saribulut, A. Teke, M. B. Latran, and M. Tümay, “Fundamentals and literature review of wavelet transform in power quality issues,” J. Electr. Electron. Eng. Res., vol. 5, no. 1, pp. 1–8, 2013. https://doi.org/10.5897/JEEER2013.0435.
- Search Google Scholar
- Export Citation
[39]↑
MATLAB & Simulink-MathWorks, “Choose a wavelet,”. [Online]. https://kr.mathworks.com/help/wavelet/gs/choose-a-wavelet.html. Accessed: Nov. 15, 2020.
- Search Google Scholar
- Export Citation
[40]↑
MATLAB & Simulink-MathWorks, “Continuous and discrete wavelet transforms,”. [Online]. https://kr.mathworks.com/help/wavelet/gs/continuous-and-discretewavelet-transforms.html. Accessed Nov. 15, 2020.
- Search Google Scholar
- Export Citation
[41]↑
L. Navarro, G. Courbebaisse, and M. Jourlin, “Logarithmic wavelets,” in Advances in Imaging and Electron Physics, vol. 183. New York, NY, USA: Academic, 2014, pp. 41–98.
- Search Google Scholar
- Export Citation
[42]↑
D. Komorowski and S. Pietraszek, “The use of continuous wavelet transform based on the fast fourier transform in the analysis of multi-channel electrogastrography recordings,” J. Med. Syst., vol. 40, no. 1, Oct. 2015. https://doi.org/10.1007/s10916-015-0358-4.
- Search Google Scholar
- Export Citation
[43]↑
J. Sadowsky, “The continuous wavelet transform: a tool for signal investigation and understanding,” Johns Hopkins APL Tech. Dig., vol. 15, no. 4, pp. 306–318, 1994.
- Search Google Scholar
- Export Citation
[44]↑
C. Liner, “An overview of wavelet transform concepts and applications,” Tech. Rep., 2010. [Online]. https://www.semanticscholar.org/paper/An-overview-of-wavelet-transform-concepts-and-Liner/52372db16936b0188f5257b80ef5804fff96c411#citing-papers.
- Search Google Scholar
- Export Citation
[45]↑
A. J. Humaidi and M. R. Hameed, “Design and performance investigation of block-backstepping algorithms for ball and arc system,” IEEE International Conference on Power, Control, Signals and Instrumentation Engineering, ICPCSI, vol. 2018, pp. 325–332, 2017.
- Search Google Scholar
- Export Citation
[46]↑
M. Al-Amidie, A. Al-Asadi, A. J. Humaidi, A. Al-Dujaili, L. Alzubaidi, L. Farhan, M. A. Fadhel, R. G. McGarvey, and N. E. Islam, “Robust spectrum sensing detector based on mimo cognitive radios with non-perfect channel gain,” Electronics, vol. 10, no. 5, 2021. https://doi.org/10.3390/electronics10050529.
- Search Google Scholar
- Export Citation
[47]↑
R. A. Kadhim, M. Q. Kadhim, H. Al-Khazraji, and A. J. Humaidi, “Bee algorithm based control design for two-links robot arm systems,” IIUM Eng. J., vol. 25, no. 2, pp. 367–380, 2024. https://doi.org/10.31436/iiumej.v25i2.3188.
- Search Google Scholar
- Export Citation
[48]↑
H. Al-Khazraji, W. Guo, and A. J. Humaidi, “Improved cuckoo search optimization for production inventory control systems,” Serbian J. Electr. Eng., vol. 21, no. 2, pp. 187–200, 2024. https://doi.org/10.2298/SJEE2402187A.
- Search Google Scholar
- Export Citation
[49]↑
H. Al-Khazraji, K. Al-Badri, R. Al-Majeez, and A.J. Humaidi, “Synergetic control design based sparrow search optimization for tracking control of driven-pendulum system,” J. Robotics Control (JRC), vol. 5, no. 5, pp. 1549–1556, 2024. https://doi.org/10.18196/jrc.v5i5.22893.
- Search Google Scholar
- Export Citation
[50]↑
F. R. Yaseen, M. Q. Kadhim, H. Al-Khazraji, and A. J. Humaidi, “Decentralized control design for heating system in multi-zone buildings based on whale optimization algorithm,” J. Européen des Systèmes Automatisés, vol. 57, no. 4, pp. 981–989, 2024. https://doi.org/10.18280/jesa.570406.
- Search Google Scholar
- Export Citation
[51]↑
H. Al-Khazraji, K. Albadri, R. Almajeez, and A. J. Humaidi, “Synergetic control-based sea lion optimization approach for position tracking control of ball and beam system,” Int. J. Robotics Control Syst., vol. 4, no. 4, pp. 1547–1560, 2024. http://doi.org/10.31763/ijrcs.v4i4.1551.
- Search Google Scholar
- Export Citation
[52]↑
A. Q. Al-Dujaili, A. J. Humaidi, Z. T. Allawi, and M. E. Sadiq, “Earthquake hazard mitigation for uncertain building systems based on adaptive synergetic control,” Appl. Syst. Innovation, vol. 6, no. 2, p. 34, 2023. https://doi.org/10.3390/asi6020034.
- Search Google Scholar
- Export Citation
[53]↑
A. Al-Dujaili, V. Cocquempot, M. E. B. E. Najjar, D. Pereira, and A. Humaidi, “Adaptive fault-tolerant control design for multi-linked two-wheel drive mobile robots. mobile robot: motion control and path planning,” in Studies in Computational Intelligence, vol. 1090, Springer International Publishing, 2023, pp. 283–329. https://doi.org/10.1007/978-3-031-26564-8_10.
- Search Google Scholar
- Export Citation
[54]↑
A. Al-Dujaili, V. Cocquempot, M. E. E. Najjar, D. Pereira, and A. Humaidi, “Fault diagnosis and fault tolerant control for n-linked two wheel drive mobile robots,” Mobile Robot: Motion Control and Path Planning. Studies in Computational Intelligence, vol. 1090, A. T. Azar, I. Kasim Ibraheem, and A. Jaleel Humaidi, Eds., Cham: Springer, 2023.
- Search Google Scholar
- Export Citation
[55]↑
M. Y. Hassan, A. J. Humaidi, and M. K. Hamza, “On the design of backstepping controller for Acrobot system based on adaptive observer,” Int. Rev. Electr. Eng., vol. 15, no. 4, pp. 328–335, 2020.
- Search Google Scholar
- Export Citation
[56]↑
W. R. Abdul-Adheem, A. T. Azar, I. K. Ibraheem, and A. J. Humaidi, “Novel active disturbance rejection control based on nested linear extended state observers,” Appl. Sci. (Switzerland), vol. 10, no. 12, p. 4069, 2020.
- Search Google Scholar
- Export Citation
[57]↑
A. F. Hasan, N. Al-Shamaa, S. S. Husain, A. J. Humaidi, and A. Al-dujaili, “Spotted hyena optimizer enhances the performance of fractional-order PD controller for tri-copter drone,” Int. Rev. Appl. Sci. Eng., vol. 15, no. 1, pp. 82–94, 2024. https://doi.org/10.1556/1848.2023.00659.
- Search Google Scholar
- Export Citation
[58]↑
S. S. Husain, A. Q. Al-Dujaili, A. A. Jaber, A. J. Humaidi, and R. S. Al-Azzawi, “Design of a robust controller based on barrier function for vehicle steer-by-wire systems,” World Electric Vehicle J., vol. 15, no. 1, p. 17, 2024. https://doi.org/10.3390/wevj15010017.
- Search Google Scholar
- Export Citation
[59]↑
A. J. Humaidi, M. R. Hameed, A. F. Hasan, A. S. M. Al-Obaidi, A. T. Azar, I. K. Ibraheem, A. Q. Al-Dujaili, A. K. Al Mhdawi, and F. A. Abdulmajeed, “Algorithmic design of block backstepping motion and stabilization control for segway mobile robot,” in Mobile Robot: Motion Control and Path Planning. Studies in Computational Intelligence, vol. 1090, A. T. Azar, I. Kasim Ibraheem, and A. Jaleel Humaidi, Eds., Cham: Springer, 2023. https://doi.org/10.1007/978-3-031-26564-8_16.
- Search Google Scholar
- Export Citation
[60]↑
A. F. Hasan, A.J. Humaidi, A. S. M. Al-Obaidi, A. T. Azar, I. K. Ibraheem, Ayad Q. Al-Dujaili, Ammar K. Al-Mhdawi, and Farah Ayad Abdulmajeed, “Fractional order extended state observer enhances the performance of controlled tri-copter UAV based on active disturbance rejection control,” in Mobile Robot: Motion Control and Path Planning. Studies in Computational Intelligence, vol. 1090, A. T. Azar, I. Kasim Ibraheem, and A. Jaleel Humaidi, Eds., Cham: Springer, 2023. https://doi.org/10.1007/978-3-031-26564-8_14.
- Search Google Scholar
- Export Citation
[61]↑
A. J. Humaidi, E. N. Tala’at, M. R. Hameed, and A. H. Hameed, “Design of adaptive observer-based backstepping control of cart-pole pendulum system,” in Proceedings of 2019 3rd IEEE International Conference on Electrical, Computer and Communication Technologies, ICECCT 2019, vol. 2019, IEEE, 2029, pp. 1–5. https://doi.org/10.1109/ICECCT.2019.8869179.
- Search Google Scholar
- Export Citation
[62]↑
A. Al-Dujaili, Y. Ma, M. El Badaoui El Najjar, and V. Cocquempot, “Actuator fault compensation in three linked 2WD mobile robots using multiple dynamic controllers,” IFAC-PapersOnLine, vol. 50, no. 1, pp. 13556–13562, 2017.
- Search Google Scholar
- Export Citation
[63]↑
Y. Ma, A. Al-Dujaili, V. Cocquempot, and M. E. Badaoui El Najjar, “An adaptive actuator failure compensation scheme for two linked 2WD mobile robots,” J. Phys. Conf. Ser., vol. 783, no. 1, 2017, Art no. 012021.
- Search Google Scholar
- Export Citation
[64]↑
A. K. Mohammed, N. K. Al-Shamaa, and A. Q. Al-Dujaili, “Super-Twisting sliding mode control of permanent magnet DC motor,” in 2022 IEEE 18th International Colloquium on Signal Processing & Applications (CSPA), 2022, pp. 347–352.
- Search Google Scholar
- Export Citation
[65]
A. Al-dujaili, V. Cocquempot, M. E. B. El Najjar, and Y. Ma, “Actuator fault compensation tracking control for multi linked 2WD mobile robots,” in 2017 25th Mediterranean Conference on Control and Automation (MED), 2017, pp. 448–453.
- Search Google Scholar
- Export Citation
[66]↑
E. K. Hamza, L. S. Mahdy, and A. A. Thabit, “Design and implementation of cognitive radio (CR) based on Xilinx FPGA,” ARPN J. Eng. Appl. Sci., vol. 14, no. 4, pp. 892–897, 2019.
- Search Google Scholar
- Export Citation
[67]↑
A. S. Mahdi Al-Obaidi, A. A. AL-Qassar, A. R. Nasser, A. Alkhayyat, A J. Humaidi, and K. I. Ibraheem, “Embedded design and implementation of mobile robot for surveillance applications,” Indonesian J. Sci. Technol., vol. 6, no. 2, pp. 427–440, 2021. https://doi.org/10.17509/ijost.v6i2.36275.
- Search Google Scholar
- Export Citation
[68]↑
E. K. Hamza, K. D. Salman, and S. N. Jaafar, “Wireless sensor network for robot navigation,” Stud. Comput. Intelligence, vol. 1090, pp. 643–670, 2023.
- Search Google Scholar
- Export Citation
[69]↑
E. K. Ibraheem and E. K. Hamza, “Load balancing performance optimization for LI-Fi/Wi-Fi HLR access points using particle swarm optimization and DL algorithms,” Int. J. Intell. Eng. Syst., vol. 15, no. 6, pp. 364–381, 2022.
- Search Google Scholar
- Export Citation
[70]↑
M. A. Fadhel, A. J. Humaidi, and S. Razzaq oleiwi, “Image processing-based diagnosis of sickle cell anemia in erythrocytes,” in 2017 Annual Conference on New Trends in Information and Communications Technology Applications (NTICT), Baghdad, Iraq, 2017, pp. 203–207. https://doi.org/10.1109/NTICT.2017.7976124.
- Search Google Scholar
- Export Citation
[71]
R. M. Mahmood, A. R. Ajel, J. K. Abed, R. A. Mahmod, A. Q. Al-Dujaili, and A. J. Humaidi, “LabVIEW-based design of smart wireless monitoring system for cardiac patients,” in 2023 IEEE 13th International Conference on System Engineering and Technology (ICSET), Shah Alam, Malaysia, 2023, pp. 46–49.
- Search Google Scholar
- Export Citation
[72]
H. Z. Khaleel and A. J. Humaidi, “Towards accuracy improvement in solution of inverse kinematic problem in redundant robot: A comparative analysis,” Int. Rev. Appl. Sci. Eng., vol. 15, no. 2, pp. 242–251, 2024. https://doi.org/10.1556/1848.2023.00722.
- Search Google Scholar
- Export Citation
[73]
R. Z. Khaleel, H. Z. Khaleel, A. Al-Hareeri, A. S. Mahdi Al-ObaidiAl-Obaidi, and A. J. Humaidi, “Improved trajectory planning of mobile robot based on pelican optimization algorithm,” J. Eur. Syst. Autom., vol. 57, no. 4, 2024. https://doi.org/10.18280/jesa.570408.
- Search Google Scholar
- Export Citation

ProCite

RefWorks

Reference Manager

BibTeX

Zotero

EndNote

Save
Cite
Email this content

Share Link

Copy this link, or click below to email it to a friend
Email this content
or copy the link directly:

https://akjournals.com/view/journals/1848/aop/article-10.1556-1848.2024.00885/article-10.1556-1848.2024.00885.xml
The link was not copied. Your current browser may not support copying via this button.

Link copied successfully

Collapse
Expand

International Review of Applied Sciences and Engineering

Print ISSN:: 2062-0810

Online ISSN:: 2063-4269

Editorial Board

Mohammad Nazir AHMAD, Institute of Visual Informatics, Universiti Kebangsaan Malaysia, Malaysia

Murat BAKIROV, Center for Materials and Lifetime Management Ltd., Moscow, Russia

Nicolae BALC, Technical University of Cluj-Napoca, Cluj-Napoca, Romania

Umberto BERARDI, Toronto Metropolitan University, Toronto, Canada

Ildikó BODNÁR, University of Debrecen, Debrecen, Hungary

Sándor BODZÁS, University of Debrecen, Debrecen, Hungary

Fatih Mehmet BOTSALI, Selçuk University, Konya, Turkey

Samuel BRUNNER, Empa Swiss Federal Laboratories for Materials Science and Technology, Dübendorf, Switzerland

István BUDAI, University of Debrecen, Debrecen, Hungary

Constantin BUNGAU, University of Oradea, Oradea, Romania

Shanshan CAI, Huazhong University of Science and Technology, Wuhan, China

Michele De CARLI, University of Padua, Padua, Italy

Robert CERNY, Czech Technical University in Prague, Prague, Czech Republic

Erdem CUCE, Recep Tayyip Erdogan University, Rize, Turkey

György CSOMÓS, University of Debrecen, Debrecen, Hungary

Tamás CSOKNYAI, Budapest University of Technology and Economics, Budapest, Hungary

Anna FORMICA, IASI National Research Council, Rome, Italy

Alexandru GACSADI, University of Oradea, Oradea, Romania

Eugen Ioan GERGELY, University of Oradea, Oradea, Romania

Janez GRUM, University of Ljubljana, Ljubljana, Slovenia

Géza HUSI, University of Debrecen, Debrecen, Hungary

Ghaleb A. HUSSEINI, American University of Sharjah, Sharjah, United Arab Emirates

Nikolay IVANOV, Peter the Great St. Petersburg Polytechnic University, St. Petersburg, Russia

Antal JÁRAI, Eötvös Loránd University, Budapest, Hungary

Gudni JÓHANNESSON, The National Energy Authority of Iceland, Reykjavik, Iceland

László KAJTÁR, Budapest University of Technology and Economics, Budapest, Hungary

Ferenc KALMÁR, University of Debrecen, Debrecen, Hungary

Tünde KALMÁR, University of Debrecen, Debrecen, Hungary

Milos KALOUSEK, Brno University of Technology, Brno, Czech Republik

Jan KOCI, Czech Technical University in Prague, Prague, Czech Republic

Vaclav KOCI, Czech Technical University in Prague, Prague, Czech Republic

Imre KOCSIS, University of Debrecen, Debrecen, Hungary

Imre KOVÁCS, University of Debrecen, Debrecen, Hungary

Angela Daniela LA ROSA, Norwegian University of Science and Technology, Trondheim, Norway

Éva LOVRA, Univeqrsity of Debrecen, Debrecen, Hungary

Elena LUCCHI, Eurac Research, Institute for Renewable Energy, Bolzano, Italy

Tamás MANKOVITS, University of Debrecen, Debrecen, Hungary

Igor MEDVED, Slovak Technical University in Bratislava, Bratislava, Slovakia

Ligia MOGA, Technical University of Cluj-Napoca, Cluj-Napoca, Romania

Marco MOLINARI, Royal Institute of Technology, Stockholm, Sweden

Henrieta MORAVCIKOVA, Slovak Academy of Sciences, Bratislava, Slovakia

Phalguni MUKHOPHADYAYA, University of Victoria, Victoria, Canada

Balázs NAGY, Budapest University of Technology and Economics, Budapest, Hungary

Husam S. NAJM, Rutgers University, New Brunswick, USA

Jozsef NYERS, Subotica Tech College of Applied Sciences, Subotica, Serbia

Bjarne W. OLESEN, Technical University of Denmark, Lyngby, Denmark

Stefan ONIGA, North University of Baia Mare, Baia Mare, Romania

Joaquim Norberto PIRES, Universidade de Coimbra, Coimbra, Portugal

László POKORÁDI, Óbuda University, Budapest, Hungary

Roman RABENSEIFER, Slovak University of Technology in Bratislava, Bratislava, Slovak Republik

Mohammad H. A. SALAH, Hashemite University, Zarqua, Jordan

Dietrich SCHMIDT, Fraunhofer Institute for Wind Energy and Energy System Technology IWES, Kassel, Germany

Lorand SZABÓ, Technical University of Cluj-Napoca, Cluj-Napoca, Romania

Csaba SZÁSZ, Technical University of Cluj-Napoca, Cluj-Napoca, Romania

Ioan SZÁVA, Transylvania University of Brasov, Brasov, Romania

Péter SZEMES, University of Debrecen, Debrecen, Hungary

Edit SZŰCS, University of Debrecen, Debrecen, Hungary

Radu TARCA, University of Oradea, Oradea, Romania

Zsolt TIBA, University of Debrecen, Debrecen, Hungary

László TÓTH, University of Debrecen, Debrecen, Hungary

László TÖRÖK, University of Debrecen, Debrecen, Hungary

Anton TRNIK, Constantine the Philosopher University in Nitra, Nitra, Slovakia

Ibrahim UZMAY, Erciyes University, Kayseri, Turkey

Andrea VALLATI, Sapienza University, Rome, Italy

Tibor VESSELÉNYI, University of Oradea, Oradea, Romania

Nalinaksh S. VYAS, Indian Institute of Technology, Kanpur, India

Deborah WHITE, The University of Adelaide, Adelaide, Australia

International Review of Applied Sciences and Engineering
Address of the institute: Faculty of Engineering, University of Debrecen
H-4028 Debrecen, Ótemető u. 2-4. Hungary
Email: irase@eng.unideb.hu

Indexing and Abstracting Services:

DOAJ
ERIH PLUS
Google Scholar
ProQuest
SCOPUS
Ulrich's Periodicals Directory

2023
Scimago
Scimago H-index	11
Scimago Journal Rank	0.249
Scimago Quartile Score	Architecture (Q2) Engineering (miscellaneous) (Q3) Environmental Engineering (Q3) Information Systems (Q4) Management Science and Operations Research (Q4) Materials Science (miscellaneous) (Q3)
Scopus
Scopus Cite Score	2.3
Scopus CIte Score Rank	Architecture (Q1) General Engineering (Q2) Materials Science (miscellaneous) (Q3) Environmental Engineering (Q3) Management Science and Operations Research (Q3) Information Systems (Q3)
Scopus SNIP	0.751

International Review of Applied Sciences and Engineering
Publication Model	Gold Open Access Online only
Submission Fee	none
Article Processing Charge	1100 EUR/article
Regional discounts on country of the funding agency	World Bank Lower-middle-income economies: 50% World Bank Low-income economies: 100%
Further Discounts	Limited number of full waivers available. Editorial Board / Advisory Board members: 50% Corresponding authors, affiliated to an EISZ member institution subscribing to the journal package of Akadémiai Kiadó: 100%
Subscription Information	Gold Open Access

International Review of Applied Sciences and Engineering
Language	English
Size	A4
Year of Foundation	2010
Volumes per Year	1
Issues per Year	3
Founder	Debreceni Egyetem
Founder's Address	H-4032 Debrecen, Hungary Egyetem tér 1
Publisher	Akadémiai Kiadó
Publisher's Address	H-1117 Budapest, Hungary 1516 Budapest, PO Box 245.
Responsible Publisher	Chief Executive Officer, Akadémiai Kiadó
ISSN	2062-0810 (Print)
ISSN	2063-4269 (Online)