A5051533976- Probabilistic and Robust Engineering Design
- Advanced Multi-Objective Optimization Algorithms
- Structural Health Monitoring Techniques
- Fault Detection and Control Systems
- Model Reduction and Neural Networks
- Advanced Aircraft Design and Technologies
- Machine Learning and Data Classification
- Manufacturing Process and Optimization
- Machine Learning and Algorithms
- Control Systems and Identification
- Stroke Rehabilitation and Recovery
- Machine Fault Diagnosis Techniques
- Advanced Bandit Algorithms Research
- Real-time simulation and control systems
- Gaussian Processes and Bayesian Inference
- Target Tracking and Data Fusion in Sensor Networks
- Computational Fluid Dynamics and Aerodynamics
- Cerebral Palsy and Movement Disorders
- Product Development and Customization
- Space Science and Extraterrestrial Life
- Systems Engineering Methodologies and Applications
- Risk and Safety Analysis
- Planetary Science and Exploration
- Neural Networks and Applications
- Aerospace and Aviation Technology
Imperial College London
2023-2024
Polytechnic University of Turin
2012-2024
Massachusetts Institute of Technology
2012-2023
Université de Poitiers
2022-2023
United Technologies Research Center
2018-2019
University of Turin
2010-2019
United Technologies Corporation (Poland)
2018
American Institute of Aeronautics and Astronautics
2013-2014
Abstract Science and Engineering applications are typically associated with expensive optimization problem to identify optimal design solutions states of the system interest. Bayesian active learning compute surrogate models through efficient adaptive sampling schemes assist accelerate this search task toward a given goal. Both those methodologies driven by specific infill/learning criteria which quantify utility respect set goal evaluating objective function for unknown combinations...
This paper proposes a data-driven strategy to assist online rapid decision making for an unmanned aerial vehicle that uses sensed data estimate its structural state, this update corresponding flight capabilities, and then dynamically replans mission accordingly. The approach comprises offline computational phases constructed address the sense–plan–act information flow while avoiding costly inference step. During phase, high-fidelity finite element simulations are used construct reduced-order...
The adoption of high-fidelity models for many-query optimization problems is majorly limited by the significant computational cost required their evaluation at every query. Multifidelity Bayesian methods (MFBO) allow to include costly responses a sub-selection queries only, and use fast lower-fidelity accelerate process. State-of-the-art rely on purely data-driven search do not explicit information about physical context. This paper acknowledges that prior knowledge domains engineering can...
A self-aware aerospace vehicle can dynamically adapt the way it performs missions by gathering information about itself and its surroundings responding intelligently. We consider specific challenge of an unmanned aerial that autonomously sense structural state re-plan mission according to estimated current health. The is achieve each these tasks in real time–executing online models exploiting dynamic data streams–while also accounting for uncertainty. Our approach combines from physics-based...
The exploration and trade-off analysis of different aerodynamic design configurations requires solving optimization problems. major bottleneck to assess the optimal is large number time-consuming evaluations high-fidelity computational fluid dynamics (CFD) models, necessary capture non-linear phenomena discontinuities that occur at higher Mach regimes. To address this limitation, we introduce an original non-myopic multifidelity Bayesian framework aimed including expensive CFD simulations...
In this paper we develop initial offline and online capabilities for a self-aware aerospace vehicle. Such vehicle can dynami- cally adapt the way it performs missions by gathering information about itself its surroundings via sensors responding in- telligently. The key challenge to enabling such is achieve tasks of dynamically autonomously sensing, planning, acting in real time. Our first steps towards achieving goal are presented here, where consider execution mapping strategies from sensed...
Abstract The multidisciplinary design optimization (MDO) of re-entry vehicles presents many challenges associated with the plurality domains that characterize problem and multi-physics interactions. Aerodynamic thermodynamic phenomena are strongly coupled relate to heat loads affect vehicle along trajectory, which drive thermal protection system (TPS). preliminary would benefit from accurate high-fidelity aerothermodynamic analysis, usually expensive computational fluid dynamic simulations....
Abstract The push toward reducing the aircraft development cycle time motivates of collaborative frameworks that enable more integrated design and their systems. ModellIng Simulation tools for Systems IntegratiON on Aircraft (MISSION) project aims to develop an modelling simulation framework. This paper focuses some recent advancements in MISSION presents a framework combines filtering process down-select feasible architectures, modeling platform simulates power system aircraft, machine...
Prognostics and health management aim to predict the remaining useful life (RUL) of a system allow timely planning replacement components, limiting need for corrective maintenance downtime equipment. A major challenge in prognostics is availability accurate physics-based representations faults dynamics. Additionally, analysis data acquired during flight operations traditionally time consuming expensive. This work proposes computational method overcome these limitations through dynamic...
Bayesian optimization is a popular framework for the of black box functions. Multifidelity methods allows to accelerate by exploiting low-fidelity representations expensive objective Popular multifidelity strategies rely on sampling policies that account immediate reward obtained evaluating function at specific input, precluding greater informative gains might be looking ahead more steps. This paper proposes Non-Myopic Optimization (NM2-BO) grasp long-term from future steps optimization. Our...
The preliminary design of a jet aircraft wing, through the use an integrated multidisciplinary environment, is presented in this paper. A framework for parametric studies wing structures has been developed on basis multilevel distributed analysis architecture with “hybrid strategy” process that able to perform deterministic optimizations and tradeoff simultaneously. particular feature proposed optimization it can different set variables, defined expressly each level, multi-level scheme using...
This work presents a modelling framework to enable comparison and trade-off study of different aircraft system architectures. The integrates computational module select feasible architectures with platform that simulates the power generation, distribution fuel consumption as well system-level models for being evaluated. Its capabilities are demonstrated case electrification primary flight control (PFCS) using electric technologies (EHA, EMA) levels ranging from conventional hydraulic...
View Video Presentation: https://doi.org/10.2514/6.2023-1092.vid Modern aerospace systems integrate a variety of multi-physical components to ensure the high-performance requirements during operational life. The increasing complexity those determines an exponential growth multiple and coupled failure modes difficult predict in advance. This hinders adoption integration new sustainable technologies, for which accurate reliable estimate incipient faults is required prevent catastrophic events....
View Video Presentation: https://doi.org/10.2514/6.2021-0894.vid Traditional methods for black box optimization require a considerable number of evaluations the objective function. This can be time consuming, impractical, and unfeasible many applications in aerospace science engineering, which rely on accurate representations expensive models to evaluate. Bayesian Optimization (BO) search global optimum by progressively (actively) learning surrogate model function along path. accelerated...