Dynamic hysteresis modelling of magnetic materials by using a neural network approach

Laudani, Antonino; Lozito, Gabriele Maria; Fulginei, Francesco Riganti

doi:10.1109/aeit.2014.7002044

Cited by 23 publications

(10 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Therefore, when the magnetic induction strength B is changed, the corresponding magnetic field strength H will be obtained. To study the magnetic properties at different frequencies, frequency f is also used as input [24], [25]. Consequently, a four-input and single-output backpropagation (BP) neural network was built to simulate the hysteresis properties of ferromagnetic materials.…”

Section: Neural Network Hysteresis Moedlmentioning

confidence: 99%

“…In recent years, researchers have paid increasing attention to the continuous development of neural network for ferromagnetic materials magnetization model due to its good nonlinear processing ability [14], [15], [16]. First, compared with the common J-A and Preisach models, the relationship between magnetic field strength H and magnetic flux density B can be directly obtained by the magnetization neural network model, thus simplifying the modeling process.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Neural Network Model for Magnetization Characteristics of Ferromagnetic Materials

Tian

Zhang

2021

IEEE Access

View full text Add to dashboard Cite

The magnetic characteristics of silicon steel sheet 30Q120 under different AC frequencies were measured by an Epstein frame in order to analyze the effects of frequency variation on the hysteresis loop of ferromagnetic materials and compare the differences of such materials at different frequencies. First, the forecasting method of the magnetic properties of ferromagnetic materials under the influence of frequency using neural network was proposed based on the measured experimental data. Hysteresis loops at different frequencies were obtained. Then, the obtained results were compared with the measured results. Second, the dynamic Jiles-Atherton hysteresis model was established based on the Jiles-Atherton hysteresis theory, and hysteresis loops at different frequencies were obtained. The accuracies of the neural network model and Jiles-Atherton hysteresis model were verified by comparing the simulation results with the measured data. Upon comparing the dynamic Jiles-Atherton hysteresis and the neural network hysteresis models, results show that the latter has better accuracy. Furthermore, the correctness and effectiveness of the proposed method are verified.

show abstract

Section: Neural Network Hysteresis Moedlmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Neural Network Model for Magnetization Characteristics of Ferromagnetic Materials

Tian

Zhang

2021

IEEE Access

View full text Add to dashboard Cite

show abstract

“…The principal scope of this research is to give a contribution to the development of the dynamic models of hysteresis for innovative soft ferromagnetic materials, exploiting artificial neural networks (ANNs). Until now, ANNs have been successfully applied in the development of both scalar [23][24][25] and vector models of static hysteresis [26][27][28], but fewer studies also take the rate dependence into account [29][30][31][32]. The main advantages of neural network-based models are related to their cheap memory allocation and high computational speed, especially when implemented at a low level of abstraction.…”

Section: Introductionmentioning

confidence: 99%

Computing Frequency-Dependent Hysteresis Loops and Dynamic Energy Losses in Soft Magnetic Alloys via Artificial Neural Networks

et al. 2022

Self Cite

View full text Add to dashboard Cite

A neural network model to predict the dynamic hysteresis loops and the energy-loss curves (i.e., the energy versus the amplitude of the magnetic induction) of soft ferromagnetic materials at different operating frequencies is proposed herein. Firstly, an innovative Fe-Si magnetic alloy, grade 35H270, is experimentally characterized via an Epstein frame in a wide range of frequencies, from 1 Hz up to 600 Hz. Parts of the dynamic hysteresis loops obtained through the experiments are involved in the training of a feedforward neural network, while the remaining ones are considered to validate the model. The training procedure is accurately designed to, firstly, identify the optimum network architecture (i.e., the number of hidden layers and the number of neurons per layer), and then, to effectively train the network. The model turns out to be capable of reproducing the magnetization processes and predicting the dynamic energy losses of the examined material in the whole range of inductions and frequencies considered. In addition, its computational and memory efficiency make the model a useful tool in the design stage of electrical machines and magnetic components.

show abstract

“…Simpler abstract models without physical background have also been proposed for hysteresis modelling as well [ 12 ]. A neural network for SMA hysteresis behavior modelling has also been applied [ 13 , 14 ]. In fact, three types of phases are observed for different thermal and mechanical combinations.…”

Section: Introductionmentioning

confidence: 99%

Assessment of SMA Electrical Resistance Change during Cyclic Stretching with Small Elongation

Sławski

Kciuk

Klein

2021

Sensors

View full text Add to dashboard Cite

In this article, changes in NiTi alloy (Flexinol) electrical resistance during cyclic stretching with small elongation were investigated. A dedicated test stand consisting of motorized vertical test stand, force gauge, and electric resistance measuring device with an accuracy of 0.006 Ω was developed. A dedicated control algorithm was developed using LabVIEW software. Changes in electrical resistance were investigated for the 0.1 mm Flexinol wire with length of 120 mm. Testing was performed in the elongation range between 0.25% and 1.5% in martensite phase. Tested samples were subjected to 30 stretching cycles with a movement speed of 10 mm/min. Obtained results show that the cyclic stretching of Flexinol wire reduces its electrical resistance with each stretching cycle. Moreover, it was noted that changes in Flexinol electrical resistance during cycling stretching depend on the assumed elongation and number of the already performed stretching cycles. The observed electrical resistance change decreases with each stretching cycle. Thus, the observed changes are greater during the first stretching cycles. For elongations exceeding 1%, the Flexinol electrical resistance in the first stretching cycle increases. In each subsequent cycle, electrical resistance decreases, as in the case of the smallest value of assumed elongation. In almost all tested cases (except in the case with 1.5% of assumed elongation), Flexinol electrical resistance after 30 stretching cycles was smaller than before the test.

show abstract

Dynamic hysteresis modelling of magnetic materials by using a neural network approach

Cited by 23 publications

References 44 publications

Neural Network Model for Magnetization Characteristics of Ferromagnetic Materials

Neural Network Model for Magnetization Characteristics of Ferromagnetic Materials

Computing Frequency-Dependent Hysteresis Loops and Dynamic Energy Losses in Soft Magnetic Alloys via Artificial Neural Networks

Assessment of SMA Electrical Resistance Change during Cyclic Stretching with Small Elongation

Contact Info

Product

Resources

About