A5080189412- Oil Spill Detection and Mitigation
- Model Reduction and Neural Networks
- Remote Sensing and LiDAR Applications
- Advanced Numerical Methods in Computational Mathematics
- Probabilistic and Robust Engineering Design
- Remote-Sensing Image Classification
- Underwater Acoustics Research
- Combustion and Detonation Processes
- Planetary Science and Exploration
- Optimal Experimental Design Methods
- Aerosol Filtration and Electrostatic Precipitation
- Numerical methods for differential equations
- Particle Dynamics in Fluid Flows
- Time Series Analysis and Forecasting
- Advanced Multi-Objective Optimization Algorithms
- Ocean Waves and Remote Sensing
- Magnetic Properties and Applications
- Synthetic Aperture Radar (SAR) Applications and Techniques
- Remote Sensing in Agriculture
- Magnetic Properties of Alloys
- Magnetic properties of thin films
- Marine and coastal ecosystems
- Maritime Navigation and Safety
- Computational Fluid Dynamics and Aerodynamics
University of Konstanz
2022
Sapienza University of Rome
2002-2020
European Space Research Institute
1999-2003
Vitrociset (Italy)
2000
A neural network approach for semi-automatic detection of oil spills in European remote sensing satellite-synthetic aperture radar (ERS-SAR) imagery is presented.The input a vector containing the values set features characterizing an spill candidate.The classification performance algorithm has been evaluated on data verified examples and look-alike.A direct analysis information content calculated also carried out through extended pruning procedure net.
A neural network approach for semi-automatic detection of oil spills in ERS-SAR imagery is presented. The input a vector containing the values set features, previously calculated by using dedicated routines, characterizing spill candidate either from point view its geometry or physical behaviour. algorithm classification performance has been evaluated on data verified examples and look-alike.
Parametric uncertainty is propagated through Reynolds-averaged Navier-Stokes (RANS) computations of a prototypical acetone/air aerosol stream flowing in dry air environment.Two parameters are considered as uncertain: the inflow velocity dissipation and coefficient that blends discrete random walk gradient-based dispersion models.A Bayesian setting employed to represent degree belief about interest terms probability theory, such described with density functions.Random variables represented by...
The paper deals with the realisation of an operational processing scheme to detect and classify ships in ERS1/2 SAR images, estimate their direction velocity. This is based on identification analysis ship wakes that are strictly related motion number orientation bright pixels corresponding ship.
Accurate parameter dependent electro-chemical numerical models for lithium-ion batteries are essential in industrial application. The exact parameters of each battery cell unknown and a process estimation is necessary to infer them. generates an accurate model able reproduce real data. field optimal input/experimental design deals with creating the experimental settings facilitating problem. Here we apply two different input algorithms that aim at maximizing observability true, parameters:...
Finite difference based micromagnetic simulations are a powerful tool for the computational investigation of magnetic structures. In this paper, we demonstrate how discretization continuous equations introduces numerical 'discretization anisotropy'. We that, in certain scenarios, anisotropy operates on an energy scale comparable to that intrinsic physical phenomena. Furthermore, illustrate selecting appropriate finite stencils and minimizing size cells effective strategies mitigate anisotropy.
In this paper the authors study a non-linear elliptic-parabolic system, which is motivated by mathematical models for lithium-ion batteries. One state satisfies parabolic reaction diffusion equation and other one an elliptic equation. The goal to determine several scalar parameters in coupled model optimal manner utilizing reliable reduced-order approach based on reduced basis (RB) method. However, states are through strongly function, makes evaluation of online-efficient error estimates...
In this paper we propose an algorithm for the bi-level optimal input design involving a parameter-dependent evolution problem. inner cycle control is fixed and parameter optimized in order to minimize cost function that measure discrepancy from some data. outer found now suitable of uncertainty parameters. The uses trust-region reduced basis approximation model with creation enrichment on-the-fly. Numerical examples illustrate efficiency proposed approach.
ABSTRACT Despite the good results obtained in a series of experiments performed jointly by satellite image providers and pollution response authorities Northern Europe, reluctance to acknowledge spaceborne data contribution oil spill surveillance is still considerable. The authors, part counter professionals, experts imagery interpretation, started collaborate with background ranging from strong doubts on suitability support control activities, substantial experience using radar satellites...