Multi-disciplinary shape optimization of an entry capsule integrated with custom neural network approximation and multi-delity approach

Claudio Cavallero, Cosimo Chiarelli, Vincenzo Mareschi, Alessio Davite, Federico Gallizio, Edmondo Minisci, Martins Sudars

Research output: Contribution to conference › Paper

Abstract

This paper describes a new integrated approach for the multi-disciplinary optimization of a entry capsule’s shape. Aerothermodynamics, Flight Mechanics and Thermal Protection System behaviour of a reference spaceship when crossing Martian atmosphere are considered, and several analytical, semi-empirical and numerical models are used. The multi-objective and multi-disciplinary optimization process implemented in Isight software environment allows finding a Pareto front of best shapes. The optimization process is integrated with a set of artificial neural networks, trained and updated by a multi-fidelity evolution control approach, to approximate the objective and constraint functions. Results obtained by means of the integrated approach with neural networks approximators are described and compared to the results obtained by a different optimization process, not using the approximators. The comparison highlights advantages and possible drawbacks of the proposed method, mainly in terms of calls to the true model and precision of the obtained Pareto front.
LanguageEnglish
Number of pages16
Publication statusPublished - 14 Sep 2011
EventEurogen 2011 Conference - Capua, Italy
Duration: 14 Sep 2011 → 16 Nov 2011

Conference

ConferenceEurogen 2011 Conference
CountryItaly
CityCapua
Period14/09/11 → 16/11/11

Fingerprint

Shape Optimization
Shape optimization
Process Optimization
Pareto Front
Neural Networks
Neural networks
Flight Mechanics
Approximation
Empirical Model
Fidelity
Atmosphere
Artificial Neural Network
Numerical models
Mechanics
Software
Optimization
Model

Keywords

  • spacecraft shape optimization
  • multi-disciplinary optimization
  • meta-modelling
  • artificial neural networks
  • Isight

Cite this

Cavallero, C., Chiarelli, C., Mareschi, V., Davite, A., Gallizio, F., Minisci, E., & Sudars, M. (2011). Multi-disciplinary shape optimization of an entry capsule integrated with custom neural network approximation and multi-delity approach. Paper presented at Eurogen 2011 Conference, Capua, Italy.
Cavallero, Claudio ; Chiarelli, Cosimo ; Mareschi, Vincenzo ; Davite, Alessio ; Gallizio, Federico ; Minisci, Edmondo ; Sudars, Martins . / Multi-disciplinary shape optimization of an entry capsule integrated with custom neural network approximation and multi-delity approach. Paper presented at Eurogen 2011 Conference, Capua, Italy.16 p.
@conference{ac3766d90fb740c8946818203cab65e0,
title = "Multi-disciplinary shape optimization of an entry capsule integrated with custom neural network approximation and multi-delity approach",
abstract = "This paper describes a new integrated approach for the multi-disciplinary optimization of a entry capsule’s shape. Aerothermodynamics, Flight Mechanics and Thermal Protection System behaviour of a reference spaceship when crossing Martian atmosphere are considered, and several analytical, semi-empirical and numerical models are used. The multi-objective and multi-disciplinary optimization process implemented in Isight software environment allows finding a Pareto front of best shapes. The optimization process is integrated with a set of artificial neural networks, trained and updated by a multi-fidelity evolution control approach, to approximate the objective and constraint functions. Results obtained by means of the integrated approach with neural networks approximators are described and compared to the results obtained by a different optimization process, not using the approximators. The comparison highlights advantages and possible drawbacks of the proposed method, mainly in terms of calls to the true model and precision of the obtained Pareto front.",
keywords = "spacecraft shape optimization , multi-disciplinary optimization , meta-modelling, artificial neural networks , Isight",
author = "Claudio Cavallero and Cosimo Chiarelli and Vincenzo Mareschi and Alessio Davite and Federico Gallizio and Edmondo Minisci and Martins Sudars",
year = "2011",
month = "9",
day = "14",
language = "English",
note = "Eurogen 2011 Conference ; Conference date: 14-09-2011 Through 16-11-2011",

}

Cavallero, C, Chiarelli, C, Mareschi, V, Davite, A, Gallizio, F, Minisci, E & Sudars, M 2011, 'Multi-disciplinary shape optimization of an entry capsule integrated with custom neural network approximation and multi-delity approach' Paper presented at Eurogen 2011 Conference, Capua, Italy, 14/09/11 - 16/11/11, .

Multi-disciplinary shape optimization of an entry capsule integrated with custom neural network approximation and multi-delity approach. / Cavallero, Claudio; Chiarelli, Cosimo; Mareschi, Vincenzo; Davite, Alessio; Gallizio, Federico; Minisci, Edmondo; Sudars, Martins .

2011. Paper presented at Eurogen 2011 Conference, Capua, Italy.

Research output: Contribution to conference › Paper

TY - CONF

T1 - Multi-disciplinary shape optimization of an entry capsule integrated with custom neural network approximation and multi-delity approach

AU - Cavallero, Claudio

AU - Chiarelli, Cosimo

AU - Mareschi, Vincenzo

AU - Davite, Alessio

AU - Gallizio, Federico

AU - Minisci, Edmondo

AU - Sudars, Martins

PY - 2011/9/14

Y1 - 2011/9/14

N2 - This paper describes a new integrated approach for the multi-disciplinary optimization of a entry capsule’s shape. Aerothermodynamics, Flight Mechanics and Thermal Protection System behaviour of a reference spaceship when crossing Martian atmosphere are considered, and several analytical, semi-empirical and numerical models are used. The multi-objective and multi-disciplinary optimization process implemented in Isight software environment allows finding a Pareto front of best shapes. The optimization process is integrated with a set of artificial neural networks, trained and updated by a multi-fidelity evolution control approach, to approximate the objective and constraint functions. Results obtained by means of the integrated approach with neural networks approximators are described and compared to the results obtained by a different optimization process, not using the approximators. The comparison highlights advantages and possible drawbacks of the proposed method, mainly in terms of calls to the true model and precision of the obtained Pareto front.

AB - This paper describes a new integrated approach for the multi-disciplinary optimization of a entry capsule’s shape. Aerothermodynamics, Flight Mechanics and Thermal Protection System behaviour of a reference spaceship when crossing Martian atmosphere are considered, and several analytical, semi-empirical and numerical models are used. The multi-objective and multi-disciplinary optimization process implemented in Isight software environment allows finding a Pareto front of best shapes. The optimization process is integrated with a set of artificial neural networks, trained and updated by a multi-fidelity evolution control approach, to approximate the objective and constraint functions. Results obtained by means of the integrated approach with neural networks approximators are described and compared to the results obtained by a different optimization process, not using the approximators. The comparison highlights advantages and possible drawbacks of the proposed method, mainly in terms of calls to the true model and precision of the obtained Pareto front.

KW - spacecraft shape optimization

KW - multi-disciplinary optimization

KW - meta-modelling

KW - artificial neural networks

KW - Isight

UR - http://www.eurogen2011.cira.it/

M3 - Paper

ER -

Cavallero C, Chiarelli C, Mareschi V, Davite A, Gallizio F, Minisci E et al. Multi-disciplinary shape optimization of an entry capsule integrated with custom neural network approximation and multi-delity approach. 2011. Paper presented at Eurogen 2011 Conference, Capua, Italy.

Access to Document