Accurate modelling of lossy SIW resonators using a neural network residual kriging approach

Angiulli, Giovanni; Carlo, Domenico De; Sgró, Annalisa; Versaci, Mario; Morabito, Francesco Carlo

doi:10.1587/elex.14.20170073

Cited by 4 publications

(1 citation statement)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the geological field, NNRK is used to predict the distribution of minerals or pollutants in soil [35][36][37][38][39]. In other fields, NNRK also has some applications [40,41]. The neural network residual Kriging algorithm is a powerful geostatistical technique that combines the strengths of neural networks and Kriging.…”

Section: Introductionmentioning

confidence: 99%

Flight Load Calculation Using Neural Network Residual Kriging

2023

View full text Add to dashboard Cite

Flight load calculation, an important step in aircraft design and optimization, typically involves millions of computations and requires significant computing resources and time. Improving the efficiency of flight load calculations while maintaining accuracy is therefore of great significance for shortening research and development cycles. This study investigated and compared multiple algorithms, including the neural network model, the Kriging surrogate model, and the neural network residual Kriging (NNRK) model, for flight load analysis. The accuracies of all models were confirmed through evaluation, with NNRK being the most efficient, making it highly suitable for flight load analysis. The flight load data of a civil aircraft, including the total weight, the center of gravity, the pitch moment of inertia, the altitude, the Mach number, the airspeed, the velocity pressure, the pitch rate, the load factor, and the angle of attack as input parameters, were used as sample data to establish models, for predicting wing loads under different flight conditions. The accuracies of all regressions were confirmed through evaluation, with NNRK being the most efficient. The flight load calculation shows that NNRK can significantly improve analysis efficiency and provide new insights into efficient and comprehensive flight load analysis.

show abstract

Section: Introductionmentioning

confidence: 99%

Flight Load Calculation Using Neural Network Residual Kriging

2023

View full text Add to dashboard Cite

show abstract

RMIL/AG: A New Class of Nonlinear Conjugate Gradient for Training Back Propagation Algorithm

Basri

Nawi

Mamat

et al. 2018

Advances in Intelligent Systems and Computing

View full text Add to dashboard Cite

Quick retrieval of effective electromagnetic metamaterial parameters by using a Multi-fidelity Surrogate Modelling approach

Angiulli

Versaci

Calcagno

et al. 2020

Eur. Phys. J. Appl. Phys.

View full text Add to dashboard Cite

Electromagnetic metamaterials (MMs) are artificial composites that exhibit exceptional physical characteristics. Their design, which relies on the retrieving of the effective medium parameters, is usually a very time-consuming process because of the high number of full-wave simulations involved in this task. To alleviate the related computational burden, we propose to use a Multi-fidelity Surrogate Modelling (MFSM) approach. Numerical results demonstrate that this methodology turns out to be promising for a quick evaluation of the scattering parameters from which the effective constitutive parameters of a MM are retrieved, as shown by two numerical examples.

show abstract

Accurate modelling of lossy SIW resonators using a neural network residual kriging approach

Cited by 4 publications

References 9 publications

Flight Load Calculation Using Neural Network Residual Kriging

Flight Load Calculation Using Neural Network Residual Kriging

RMIL/AG: A New Class of Nonlinear Conjugate Gradient for Training Back Propagation Algorithm

Quick retrieval of effective electromagnetic metamaterial parameters by using a Multi-fidelity Surrogate Modelling approach

Contact Info

Product

Resources

About