International Journal of Applied Electromagnetics and Mechanics - Volume 73, issue 4

Authors: Mognaschi, Maria Evelina

Article Type: Editorial

DOI: 10.3233/JAE-239003

Citation: International Journal of Applied Electromagnetics and Mechanics, vol. 73, no. 4, pp. 235-236, 2023

Authors: Lowther, David A.

Article Type: Review Article

Abstract: Designing an electromagnetic device, as with many other devices, is an inverse problem. The issue is that the performance and some constraints on the inputs are provided but the solution to the design problem is non-unique. Additionally, conventionally, at the start of the design process, the information on potential solutions needs to be generated quickly so that a designer can make effective decisions before moving on to detailed performance analysis, but the amount of information that can be obtained from simple analysis tools is limited. Machine learning may be able to assist by increasing the amount of information available at …the early stages of the design process. This is not a new concept, in fact it has been considered for several decades but has always been limited by the computational power available. Recent advances in machine learning might allow for the creation of a more effective “sizing” stage of the design process, thus reducing the cost of generating a final design. The goal of this paper is to review some of the work in applying artificial intelligence to the design and analysis of electromagnetic devices and to discuss what might be possible by considering some examples of the use of machine learning in several tools used in conventional design, which have been considered over the past three decades. Show more

Keywords: Electrical machine design, machine learning, key performance indicators, material properties, optimization

DOI: 10.3233/JAE-230104

Citation: International Journal of Applied Electromagnetics and Mechanics, vol. 73, no. 4, pp. 237-254, 2023

Authors: Mikami, Ryosuke | Sato, Hayaho | Hayashi, Shogo | Igarashi, Hajime

Article Type: Research Article

Abstract: This paper proposes a multi-objective optimization method for permanent magnet motors using a fast optimization algorithm, Covariance Matrix Adaptation Evolution Strategy (CMA-ES), and deep learning. Multi-objective optimization with topology optimization is effective in the design of permanent magnet motors. Although CMA-ES needs fewer population size than genetic algorithm for single objective problems, this is not evident for multi-objective problems. For this reason, the proposed method generates training data by solving the single-objective optimization multiple times using CMA-ES, and constructs a deep neural network (NN) based on the data to predict performance from motor images at high speed. The deep NN …is then used for fast solution of multi-objective optimization problems. Numerical examples demonstrate the effectiveness of the proposed method. Show more

Keywords: Deep learning, CNN, multi-objective optimization, CMA-ES, NSGA-II, PM motor

DOI: 10.3233/JAE-230077

Citation: International Journal of Applied Electromagnetics and Mechanics, vol. 73, no. 4, pp. 255-264, 2023

Authors: Rao, Shaowei | Yang, Shiyou | Tucci, Mauro | Barmada, Sami

Article Type: Research Article

Abstract: In this contribution a methodology to diagnose transformer faults based on Dissolved Gas Analysis (DGA) by using a convolutional neural network (CNN) is proposed. The algorithm to transform the gas contents (resulting from the DGA analysis) into feature maps is introduced, and the resulting feature maps are the input of the CNN. In order to take into account the fact that the data set is imbalanced, the improved Synthetic Minority Over-Sampling Technique (SMOTE) is combined with the data cleaning technique to protect the CNN from training bias. The effect of the CNN architecture on the classification performance is also investigated …to determine the optimal CNN parameters. All the above mentioned possibilities are tested and their performance investigated; in addition, a final test on the IEC TC 10 transformer fault database validates the accuracy and the generalization potential of the proposed methodology. Show more

Keywords: Convolutional neural network, deep learning, DGA, fault diagnosis, SMOTE, transformer

DOI: 10.3233/JAE-230011

Citation: International Journal of Applied Electromagnetics and Mechanics, vol. 73, no. 4, pp. 265-281, 2023

Authors: Jiang, Hao | Zhang, Hongwei | Chen, Jing | Xiao, Sa | Miao, Xiren | Lin, Weiqing

Article Type: Research Article

Abstract: The top oil temperature in ultra-high voltage (UHV) reactors has attracted enormous interest due to its wide applications in fault diagnosis and insulation evaluation. In this work, the precise prediction method based on the Seq2Seq module with the convolutional block attention mechanism is proposed for the UHV reactor. To reduce the influence of vibratility and improve computational efficiency, a combination of the encoding layer and decoding layer named Seq2Seq is performed to reconstruct the complex raw data. The convolutional block attention mechanism (CBAM), composed of spatial attention and channel attention, is utilized to maximize the use of information in data. …The Seq2Seq-CBAM is established to forecast the variation tendency of the oil temperatures in the UHV reactor. The experimental results show that the proposed method achieves high prediction accuracy for the top oil temperature in both single-step and multi-step. Show more

Keywords: UHV reactor, top oil temperature, attention, convolution block attention mechanism (CBAM), online detection scenario

DOI: 10.3233/JAE-230022

Citation: International Journal of Applied Electromagnetics and Mechanics, vol. 73, no. 4, pp. 283-302, 2023

Authors: Di Barba, Paolo | Januszkiewicz, Łukasz

Article Type: Research Article

Abstract: In modern wireless telecommunication systems, antenna arrays are widely used as elements of multiple – input multiple – output technology. In the fifth-generation systems, arrays are utilized to realize beamforming that forms the radiation pattern of the base station in the direction of the mobile user. This requires the utilization of many-element antenna arrays that are precisely controlled to achieve the required radiation properties. In this paper we apply the concept of deep neural network to model antenna array radiation properties. In this proof-of-concept research we aim at investigating to what extent it is possible to use deep neural networks …for modeling antenna arrays. We consider a full-wave model of linear array with a reflector, which was controlled by the phase and amplitude of the signals feeding the elementary radiators. The applied method made it possible to solve the direct and inverse problems. The results that we obtained show that deep neural networks are able to model antenna array properties. Show more

Keywords: Deep neural networks, antenna array, radiation pattern, method of moments

DOI: 10.3233/JAE-230086

Citation: International Journal of Applied Electromagnetics and Mechanics, vol. 73, no. 4, pp. 303-320, 2023

Authors: Hizem, Moez | Ben Saada, Aymen | Ben Mbarek, Sofiane | Choubani, Fethi

Article Type: Research Article

Abstract: Human-Like digital models have been around for quite some time. They significantly contributed to the increase of the accuracy of the whole-body-average specific absorption rate estimations. However, the anatomical and morphological diversity between human beings has not yet been embraced by the actual anthropomorphic models for several reasons such as financial costs, excessive exposure of volunteers to electromagnetic waves, and the required number of technical experts needed to build one voxelized model. Recently, machine learning has been used to reduce the complexity of certain tasks. Yet, at least, having an anthropomorphic model per nation is still far away to achieve. …To reduce the building cost of new human-like models, we build on the success of anthropomorphic models and machine learning to derive mathematical equations that make it possible to predict the Whole-body-average SAR from low frequencies up to twice the resonance frequency without any cost and excessive electromagnetic exposure of new volunteers. The completely new machine learning based equations are applicable for any age, ethnic group, and for both genders. They depend only on the human body’s morphological (height and weight) and anatomical parameters (tissue weights). In this work, we first address the whole-body-average SAR peak and we present a set of two estimators. In second, we show that the resonance frequency is not only a function of the height of the human body, to end up with a third estimation for the resonance frequency. These completely new estimators are finally combined into a novel equation that links the whole-body-average SAR to the frequency. It shows the accurate prediction for low frequencies (10 MHz) up to twice the resonance frequency. The derived estimators for the maximum WBASAR and the resonance frequencies showed better results for low frequency exposure. Show more

Keywords: SAR, exposure, frequency, voxel, anthropomorphic, dosimetry

DOI: 10.3233/JAE-230025

Citation: International Journal of Applied Electromagnetics and Mechanics, vol. 73, no. 4, pp. 321-338, 2023

Authors: Kłosowski, Grzegorz | Rymarczyk, Tomasz | Wójcik, Dariusz

Article Type: Research Article

Abstract: The main problem with any tomography is the transformation of measurements into images. It is the so-called “inverse problem”, which, due to its indeterminacy, can never be solved perfectly. An additional factor contributing to the deterioration of the quality of tomograms is measurement noise. This article shows how to denoise electrical capacitance tomography measurements using the LSTM autoencoder. The presented model is two-staged. First, the autoencoder is trained using very noisy measurements. Then, the decoder autoencoder generates a training set to using activations ofe the latent layer. In the second stage, the LSTM network is trained, which has encoder latent …layer activations at the input and pattern images at the output. The results of the experiments show that using an autoencoder to denoise the measurements improves the reconstruction quality. Show more

Keywords: Electrical tomography, industrial tomography, inverse problem, LSTM networks, autoencoders

DOI: 10.3233/JAE-230013

Citation: International Journal of Applied Electromagnetics and Mechanics, vol. 73, no. 4, pp. 339-352, 2023

Authors: Rymarczyk, Tomasz | Kulisz, Monika | Kłosowski, Grzegorz

Article Type: Research Article

Abstract: This study concerns research on using electrical impedance tomography (EIT) to image moisture inside the porous walls of buildings. In order to transform the electrical measurements into the values of the reconstructed 3D images, a neural network containing the LSTM layer was used. The objective of the study was to evaluate the impact of various loss functions on the efficacy of a neural network’s learning process. During the training process, three distinct variations of the loss function were employed, namely mean squared error (MSE), Huber, and a hybrid of MSE + Huber, to attain the desired outcome. Given that the …primary focus of the study was on the loss function, the particular neural network architecture employed was deemed non-essential. In order to minimize the influence of the neural network architecture on the outcomes of the test, a comparatively uncomplicated neural model was implemented, comprising a solitary LSTM layer and a single fully connected layer. Show more

Keywords: Machine learning, neural networks, electrical tomography, moisture inspection

DOI: 10.3233/JAE-230083

Citation: International Journal of Applied Electromagnetics and Mechanics, vol. 73, no. 4, pp. 353-367, 2023

Authors: Shi, Jinpeng | Wang, Donglai | Zhao, Yan | Li, Chengze | Zhang, Aijun

Article Type: Research Article

Abstract: The radiation of adjacent field sources has a specific spatial correlation. In order to suppress electromagnetic disturbance and improve the electromagnetic compatibility of secondary equipment, the electric field’s spatial coupling characteristics and distribution law should be mastered. Therefore, a method for predicting the spatial electric field generated by substation switching operation based on the Atomic Orbital Search-Graph Convolution Network- Long and Short-Term Memory (AOS-GCN-LSTM) model is presented to deal with this problem. First, the GCN is used to construct graph data according to node characteristics and topology information. The feature selection uses the Maximum Information Coefficient (MIC) to extract the …spatial correlation of the adjacent field source radiation. At the same time, the LSTM is used to capture the temporal correlation characteristics of different position field strengths in space. Then, the AOS is used to optimize the model with a hyperparameter. In addition, the simulation data of the full-wave simulation model of the spatial electric field generated by switch operation in a 220 kV GIS substation is an example of verification. The results show that the prediction error of the proposed method is below 3%, and it has strong adaptability to the application environment and good prediction performance. Show more

Keywords: Switch operation, spatial electric field, AOS-GCN-LSTM model, full-wave simulation, field strength prediction

DOI: 10.3233/JAE-230089

Citation: International Journal of Applied Electromagnetics and Mechanics, vol. 73, no. 4, pp. 369-387, 2023

Authors: Di Barba, Paolo | Dughiero, Fabrizio | Forzan, Michele | Lowther, David A. | Marconi, Antonio | Mognaschi, Maria Evelina | Sykulski, Jan K.

Article Type: Research Article

Abstract: The authors explore the possibility of applying a convolutional Naeural Network (CNN) to the solution of coupled electromagnetic and thermal problem, focusing on the classical problem of induction heating systems, traditionally solved by resorting to Finite Element (FE) models. In fact, FE modelling is widely used in the design of induction heating systems due its accuracy, even if the solution of a coupled nonlinear problem is expensive in terms of computational time and hardware resources, notably in 3D analysis. A model based on CNN could be an interesting alternative; in fact, CNN is a learning model selected for its …excellent ability to converge, even when trained with a limited dataset. CNNs are able to treat images as input and they are used here as follows: given a temperature map in the workpiece, identify the corresponding vector of current, frequency and process heating time; this mapping is a model of the inverse induction heating problem. Specifically, we consider as an example the induction heating of a cylindrical steel billet, made of C45 steel, placed in a solenoidal inductor coil exhibiting the same axial length of the billet (TEAM 36 problem). A thorough heating process is usually applied before hot working of the billet, as in an extrusion process, but this methodology can be applied also in the design of induction hardening processes. First, a CNN has been trained from scratch by means of a dataset of FE solutions of coupled electromagnetic and thermal problems. For the sake of a comparison, a transfer learning technique is applied using GoogLeNet, i.e. a Deep Convolutional Neural Network able to classify images: starting from the pre-trained GoogLeNet, its training has been subsequently refined with the dataset of solutions from FE analyses. When the training dataset contains a limited number of samples, GoogleNet shows good accuracy in predicting the process parameters; in the case of a high number of samples in the training set, namely beyond a threshold like e.g. 1500, both CNNs show good accuracy of the result. Show more

Keywords: Numerical modelling, coupled fields, neural network, induction heating, finite-element analysis

DOI: 10.3233/JAE-230087

Citation: International Journal of Applied Electromagnetics and Mechanics, vol. 73, no. 4, pp. 389-398, 2023

Authors: Ma, Yangyang | Li, Yongjian | Chen, Ruiying | Yue, Shuaichao | Sun, He

Article Type: Research Article

Abstract: With the increase in power electronic equipment in power system, the excitation of ferromagnetic materials often involves a large number of harmonics. Therefore, it is necessary to construct an accurate dynamic hysteresis model to adapt to this complicated operating state of electrical equipment. In this paper, a Hybrid Dynamic Hysteresis Model (HDHM), which can effectively characterize the harmonic excitation of materials is studied based on the Preisach model and Stacked Auto-Encoder (SAE) model. The static part of this model takes the form of the inverse Preisach model. And the Multiple Dynamic Hysteresis Model Set (MDHMS) is constructed by multiple dynamic …models of eddy currents and excess characteristics of the ferromagnetic materials. The dynamic part of the HDHM takes the form of the model structure combining the Stacked Auto-encoder and the MDHMS. The calculation results of the hysteresis loop and ferromagnetic loss in the harmonic condition of silicon steel sheet proves the validity of this model. Moreover, compared with the conventional dynamic hysteresis model, the HDHM has better accuracy and generalization ability. Show more

Keywords: Hybrid dynamic hysteresis model, the Preisach model, multiple-model set theory, ferromagnetic materials, iron loss

DOI: 10.3233/JAE-220112

Citation: International Journal of Applied Electromagnetics and Mechanics, vol. 73, no. 4, pp. 399-413, 2023

Authors: Mušeljić, Eniz | Roppert, Klaus | Domenig, Lukas Daniel | Köstinger, Alice Reinbacher | Kaltenbacher, Manfred

Article Type: Research Article

Abstract: This paper is about the parameter identification of an energy based hysteresis model from measurements by employing automatic differentiation and neural networks. We first introduce the energy based hysteresis model and the parameters which are to be identified. Then we show how the model can benefit from automatic differentiation. After that we incorporate a parametrization of the energy based hysteresis model via distribution functions and identify the parameters of the distribution function. Then, the hysteresis model is sampled and the generated datasets are used to train neural networks to predict the hysteresis parameters. The described methods are tested and verified …on synthetic as well as measurement data. Show more

Keywords: Optimization; parameter identification; hysteresis; machine learning; neural networks

DOI: 10.3233/JAE-230107

Citation: International Journal of Applied Electromagnetics and Mechanics, vol. 73, no. 4, pp. 415-427, 2023