CUED Publications database

Review of improved Monte Carlo methods in uncertainty-based design optimization for aerospace vehicles

Hu, X and Chen, X and Parks, GT and Yao, W (2016) Review of improved Monte Carlo methods in uncertainty-based design optimization for aerospace vehicles. Progress in Aerospace Sciences, 86. pp. 20-27. ISSN 0376-0421

Full text not available from this repository.

Abstract

© 2016 Elsevier Ltd Ever-increasing demands of uncertainty-based design, analysis, and optimization in aerospace vehicles motivate the development of Monte Carlo methods with wide adaptability and high accuracy. This paper presents a comprehensive review of typical improved Monte Carlo methods and summarizes their characteristics to aid the uncertainty-based multidisciplinary design optimization (UMDO). Among them, Bayesian inference aims to tackle the problems with the availability of prior information like measurement data. Importance sampling (IS) settles the inconvenient sampling and difficult propagation through the incorporation of an intermediate importance distribution or sequential distributions. Optimized Latin hypercube sampling (OLHS) is a stratified sampling approach to achieving better space-filling and non-collapsing characteristics. Meta-modeling approximation based on Monte Carlo saves the computational cost by using cheap meta-models for the output response. All the reviewed methods are illustrated by corresponding aerospace applications, which are compared to show their techniques and usefulness in UMDO, thus providing a beneficial reference for future theoretical and applied research.

Item Type: Article
Subjects: UNSPECIFIED
Divisions: Div A > Energy
Depositing User: Cron Job
Date Deposited: 17 Jul 2017 19:00
Last Modified: 18 Nov 2017 21:39
DOI: