232 results
Filters
Settings
Now showing1 - 10 of 232
- Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Prediction of frequency-domain data for EMC of an electric vehicle powertrain with Gaussian processes and ANN-based embeddingFor the electromagnetic compatibility (EMC) of an electric vehicle (EV), investigation of the frequency-domain behavior of the powertrain and its components is critical to ensure compliant, robust and cost-effective design. Time-consuming and expensive measurements and simulations are common to obtain the frequency-domain data, like EMC filter transfer functions and emissions. Machine learning (ML) techniques can be used to build surrogate models with very fast evaluation speed, though these models lose accuracy when training data is scarce regarding design-parameter variations. This issue is aggravated by the fact that the frequency axis is typically densely sampled which can quickly overload data-efficient ML-techniques. Here, an embedding of the frequency-domain data using artificial neural nets in an autoencoder configuration is used to reduce the number of data points that need to be handled. This allows for the usage of the data-efficient Gaussian process regression algorithm to perform the predictions over the EMC design space on the embedded data. The method is demonstrated with two EMC tasks showing high predictive accuracy and efficiency. First, the prediction of the insertion loss of an EMC filter for the high-voltage powertrain of an EV with data obtained from 3D simulation. Second, the prediction of conducted emissions of the high-voltage powertrain of an EV with data obtained from a transient simulation.Publicationtype: Conference Paper25 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Modeling and improving transmission of high-speed vias on PCBs up to 200 Ghz(IEEE, 2026-06-14); ; ; ; Sismanis, NikolaosAs high-speed interconnects on printed circuit boards (PCBs) advance into the beyond 400 Gbps domain, accurate characterization of vias becomes critical. Via arrays are investigated at frequencies up to 200 GHz using different full-wave (FW) simulation tools, yielding good correlation. Scattering phenomena play a more significant role than losses in the behavior of these arrays. Considerations essential for improved interconnect performance in next-generation highspeed communication systems, increasing bandwidth by 20 GHz, are outlined. Coax vias and interstitial ground vias improving transmission are investigated.Publicationtype: Conference Paper3 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Using decision trees for fast error identification in physics-based modeling of PCB structuresPhysics-based (PB) via models for printed circuit board (PCB) simulations are computationally efficient and have proven accurate up to 40 GHz. However, they require validation against time-intensive full-wave (FW) electromagnetic solvers when applied to faster links at 100 GHz. This study presents an error identification approach using a decision tree (DT) that categorizes designs into low- and high-error classes based on PCB design parameters, enabling targeted FW validation only where needed. The DT's interpretability reveals which geometric and material parameters most influence PB model accuracy.Publicationtype: Conference Paper10 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Behavior of crosstalk in environmentally-impacted automotive PCB interconnects and effect on signal transmission(IEEE, 2026-06); ; ; ; In the automotive industry, interconnect design is impacted by the operational conditions (temperature and humidity) encountered in practice. This paper extends the use of environmentally-dependent dielectric and conductor models, derived in our previous work, to evaluate key metrics for simulated crosstalk-impacted links operating under practical vehicle conditions up to 40 GHz. These metrics include power sum near- and far-end crosstalk (PSNEXT and PSFEXT), and the interconnect crosstalk ratio (ICR). Herein, an analysis of the dependency on temperature and moisture content of crosstalk metrics is presented for two interconnect configurations based on single-ended and differential signaling. Initial observations for PSNEXT and ICR show negligible variation under harsh environmental conditions while PSFEXT shows a considerable reduction as frequency increases.Publicationtype: Conference Paper6 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Systematic and fast crosstalk modeling of 10Base-T1S data links in the aircraft cabinBase-T1S enables multidrop Single Pair Ethernet links in aircraft cabins where adjacent unshielded twisted pairs carrying AC power or data signals may be routed in parallel. Crosstalk for identical twist rates under installation-specific geometric variation is evaluated with a computationally efficient modeling methodology based on 2D cable and SPICE simulations, covering 100 Hz to 200 MHz in time and frequency domains. Geometric parameters are swept to assess their influence on crosstalk. For a representative use case worst-case crosstalk is evaluated and model predictions are compared against measurements. The results indicate that the 2D-based workflow is sufficient for the considered aircraft-cabin use case.Publicationtype: Conference Paper17 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Identification of critical frequencies in wideband near-field spatial scans(IEEE, 2026-03-31); ; ; ; Identifying critical frequencies of a device under test (DUT) from its near-field scans is an essential step in integrating near-field scanning into electromagnetic compatibility (EMC) testing. Indeed, real electronic devices may exhibit radiated emissions at multiple frequencies, which are unknown a priori. To this end, three computational methods are introduced to identify the frequencies at which the DUT is radiating: the space-filled spectrogram analysis, the grey level co-occurrence matrix approach, and spatial fast Fourier transform approach. Although a single general solution is not provided, each method leverages different properties of the measurements to extract significant features and identify frequencies of interest. Therefore, the choice of which method to apply depends on the characteristics of the DUT and on the requirements of the specific measurement campaign. The proposed methods are evaluated on a common dataset, assessing their computational efficiency, sensitivity to patterns, and robustness against noise. Case studies demonstrate their successful application in real-world scenarios and their comparative analysis provides insights into selecting the most appropriate approach for a specific application.Publicationtype: Journal Article9 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Methods of machine learning for analysis and decoupling of power delivery networks on printed circuit boardsIncreasing challenges in the design process of power delivery networks on printed circuit boards require new and adapted tools. Therefore, in this thesis artificial neural networks are investigated to enhance the design process. Challenges and their possible solutions are discussed. For the data driven investigations more than 100000 numerical simulations of printed circuit board variations using a physics-based modeling approach are performed. Additionally, more than 124000 decoupling capacitor (decap) terminations on the power delivery networks are processed. Validations of some printed circuit board variations against a ommercial full-wave finite element method solver are included. The investigations resulted in the development of the publicly available SI/PI-Database. The database holds a majority of the printed circuit board descriptions and simulation results. It helps to increase the reusability of once created data samples for different investigations throughout this and future work. Here, the database is used in combination with a fine-tuning approach of the artificial neural network training process to increase the data efficiency of once created data samples. In the fine-tuning process the existing data samples are used to better initialize the artificial neural network for the trainingPublicationtype: Doctoral Thesis55 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: with files , Development of data-efficient machine learning approaches for the modeling and design in electromagnetic compatibility and radio-frequency engineering(2026); This thesis explores the use of machine learning (ML) in electromagnetic compatibility (EMC) engineering and related fields, focusing on common challenges such as model interpretability, high-dimensional design spaces, and data acquisition costs. It investigates ML methods that incorporate physical insights to improve modeling accuracy and reliability. Strategies to enhance data-efficiency and promote dataset reuse are proposed to support reproducibility and collaboration. Recommendations for future research and community efforts are provided to advance ML adoption in EMC and related fields. The results demonstrate the potential of ML to accelerate the design and modeling processes in EMC.Publicationtype: Doctoral Thesis31 19 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: with files , Optimization of modal reflection parameters of PCB connector and common-mode choke using Gaussian process regression and segmentation(Institute of Electrical and Electronics Engineers, 2026); ; ; ; In this contribution the electrical properties of a high-speed signal transition from a connector onto a printed circuit board (PCB) followed by a common-mode choke (CMC) in a surface mount package in terms of its modal, i.e., differentialand common-mode, reflection parameters has been investigated. An implementation of segmentation, i.e. domain decomposition and subsequent composition of subdomains, in combination with Gaussian process regression (GPR) to optimize the reflections of the PCB connectors and CMC integration on the board is presented and analyzed. GPR is used with Bayesian optimization to find an optimal set of design parameters for a 3D electromagnetic model developed in a full-wave electromagnetic field solver up to 20 GHz. The results indicate that an engineeringinformed domain decomposition strategy can reduce the computational burden of GPR-based optimization by lowering the per-evaluation full-wave cost. This surrogate-based optimization approach reduces the number of required full-wave simulations, enabling more efficient use of computational resources.Publicationtype: Conference Paper70 43 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Developments in machine learning based modeling and design for electromagnetic compatibility(IEEE, 2025-12-24); ; ; ; Machine learning (ML) has gained traction and emerged as one of most researched topics in recent years. Many engineering disciplines are taking advantage of a multitude of developed ML tools to enhance the engineering workflow, solve complicated problems or speedup design processes. This paper presents an overview on past and present developments of ML-based modeling and design in the field of electromagnetic compatibility (EMC) and the related fields of signal integrity (SI), power integrity (PI), bioelectromagnetics (BioEM) and electromagnetic field (EMF) scans. A critical analysis of recent research on ML applications is carried out to identify the arising opportunities and challenges facing the wide adoption of ML for the modeling and design in EMC. Based on current trends, a vision for ML-based modeling and design in EMC is presented in an attempt to foresee the direction of near-future research. A set of recommendations is formulated to tackle actual challenges facing a wider adoption. ML integration in EMC engineering has a great potential to revolutionize the modeling and design processes, despite the open challenges and the rising complexities of the faced problems.Publicationtype: Journal Article78