Conference Agenda

This work presents a wavelet regularization for electrical resistance tomography applied to thin conductive materials. The proposed method, implemented using the fast iterative shrinkage thresholding algorithm, is successfully applied to test measurements of a fluorinated tin oxide film with a known defect and compared with the results obtained by a traditional Tikhonov regularization.

PCB-based Inductive Position Sensors (IPS) are advantageous with respect to resistive, capacitive or Hall-effect alternatives since they allow for contactless sensors that are immune to stray fields and robust to temperature changes, all while being easy to manufacture by using conventional PCB fabrication techniques.

In this work we present a novel IPS design for measuring the angle of a target conductor with respect to the sensing coils, and the distance between them. Simulations are used to build a dataset for training a neural network to solve the inverse problem of estimating tilt and airgap from the measured voltages. The trained network is tested against experimental data measured from a physical prototype.

In this paper, an improved method for accurately estimating the length and depth of defects using magnetic flux leakage (MFL) and eddy current (EC) signals is proposed. While detailed shapes of defects influence the magnitude of MFL signal, the signal appears blunt shape, making it difficult to determine the length and depth of defects. Magnetization-based eddy current testing (MB-ECT) can detect local permeability perturbation caused by defects, enabling the characterization of the detailed features of defects. By combining the MFL and EC signals, length and depth estimation performance can be improved. The parameters of defect such as defect opening, sharpness, and the position of maximum depth are estimated by combining the MFL and EC signals. The defect shapes are represented in the function of these parameters and improved algorithm for sizing depth is proposed. In 2-page digest, MFL and EC signals are analyzed using three defect shapes, and ideas for estimating the detailed features of defects are deduced. Detailed algorithm will be presented in full paper with further simulations and experimental results.

This study proposes an algorithm using magnetic sensors to robustly detect the position and orientation of a magnetic capsule endoscope (MCE) under the external magnetic field (EMF) generated by a magnetic navigation system (MNS). The magnetic flux density generated by the MCE was separated from the flux density data measured by the magnetic sensor array, and a magnetic field model was constructed using linear interpolation to overcome the limitations of the point-dipole model. Additionally, a weighting factor based on the signal-to-noise ratio (SNR) of each sensor was introduced to enable accurate detection of position and orientation even in noisy environments. Experiments were conducted to confirm that the proposed algorithm enhances the detection performance of the MCE.

This study presents a computational and experimental verification of a measurement system with slotted ring specimens for electromagnetically characterizing uncompressed, i.e., weak magnetic powders, which present promising materials for modern electric drives. Metrological challenges in characterizing feebly magnetic bulk materials, such as the stray flux compensation, are addressed. Therefore, a new signal model is analytically derived and validated using 3D finite element analysis, ensuring the accurate indirect determination of the specimen's averaged field quantities. Additionally, the novel setup and the introduced signal model are verified with the help of 12 samples made of different iron powders and a detailed comparison of their magnetization characteristics with conventional steel sheets to demonstrate the system's ability to acquire the specimen's electromagnetic behavior.

Sensorless rotor angle estimation methods are becoming increasingly important due to the growing demands on electric motors. However, their application is often constrained by insufficient rotor angle precision and additional energy loss. This paper addresses these challenges under dynamic operating conditions by presenting a method to predict the optimal operating range of a permanent magnet synchronous motor with respect to injection frequency and loads when using zero-sequence voltage with a high-frequency injection method for rotor angle detection. The approach is based on a finite element analysis of inductance variations across different loads over one electrical rotation in frequency domain.

We propose a fast method to characterize inductive position sensors with zero physical prototypes. The technique is based on an in-house developed electromagnetic simulation tool which shows three orders of magnitude improvement with respect to the most widely used commercial software. This simulation software is used for producing synthetic data for training a machine learning surrogate model based on a neural network. In this way, the characterization of a sensor takes just milliseconds. This opens the possibility of devising a Design Support System (DesSS), a software that guide the user in sensor design for the best possible outcome, thereby saving development time.

The accurate characterization of local magnetic properties in electrical steel sheets is critical for understanding material behavior and optimizing the performance in electrical machines. This study focuses on determining the local anhysteretic magnetic behavior near cutting edges, where material degradation significantly influences its material parameters. A high-precision measurement system is utilized to detect magnetic flux leakage (MFL) at cutting edges, followed by the application of an inverse scheme to compute the local magnetic material properties based on measured and simulated data. Results from a numerical case study demonstrate the feasibility of the proposed methodology.

Magnetometers mounted on large magnetic carriers to monitor the target external magnetic field are affected by interfering magnetic fields generated by the carrier body, resulting in deviations from the true target value. To address this, a compensation process based on the pre-assessment of interfering fields is typically required to achieve accurate target-field estimation. However, inaccuracies in the evaluation of interference fields or operational changes in the carrier’s geometry and magnetic properties can lead to residual errors that existing compensation methods cannot effectively mitigate. This paper proposes an optimal design method for an interference suppression core (ISC) to overcome such challenges. The ISC is optimized to produce an induced magnetic field that passively suppresses the interference field from the carrier body while maintaining the integrity of the target external field. The optimization method employs continuum sensitivity analysis to derive the design velocity of the ISC, ensuring compliance with the multiple design objectives. A design example validates the effectiveness of the proposed method, demonstrating the ISC’s potential to reduce errors in magnetic field measurement systems.

This paper proposes a capacitive non-destructive testing (NDT) technique for detecting surface defects in pipeline inspection. Finite Element Analysis (FEA) simulations were conducted to analyze changes in surface charge density, electric field and capacitance caused by defects. Experimental data for various defect sizes were obtained using the developed capacitive sensor, and the results were compared with the simulations. Consequently, this confirms that the principle of capacitive sensors is suitable for defect measurement. This research is expected to provide a reference for research in NDT and capacitive sensor applications.

This study explores the application of Machine Learning (ML) and Artificial Intelligence (AI) techniques to enhance non-destructive eddy current testing (ECT) for detecting and characterizing defects in conductive materials. COMSOL Multiphysics is employed to model the eddy current sensor and generate a comprehensive database of simulation data, representing various defect scenarios with different sizes and geometries. This dataset is used to train ML models designed to accurately predict defect dimensions and locations. The integration of ML and AI into ECT shows a significant improvement in the accuracy of defect size estimation compared to traditional methods, underscoring the potential of these advanced techniques to improve the reliability and efficiency of non-destructive evaluation in industrial applications

This paper present the modeling of the diffusion layer thickness variation impact on eddy currents signals. The numerical modeling of laminate with the diffusion layer is realized using finite elements code developed under Matlab environment. The solving of the 2D electromagnetic equation, in (r,z) plane, is performed by considering magnetic permeability variation of diffusion layer. A comparison between the simulation results and experimental signals shows a good agreement.