Surrogate models for predicting noise emission and aerodynamic performance of propellers

Created by W.Langdon from gp-bibliography.bib Revision:1.8010

@Article{POGGI:2021:AST,

• author = "Caterina Poggi and Monica Rossetti and Giovanni Bernardini and Umberto Iemma and Cristiano Andolfi and Christian Milano and Massimo Gennaretti",
• title = "Surrogate models for predicting noise emission and aerodynamic performance of propellers",
• journal = "Aerospace Science and Technology",
• year = "2021",
• volume = "125",
• pages = "107016",
• month = jun,
• keywords = "genetic algorithms, genetic programming, Propeller aerodynamics and aeroacoustics, Surrogate models, Artificial Neural Network, ANN",
• ISSN = "1270-9638",
• URL = "https://ui.adsabs.harvard.edu/abs/2022AeST..12507016P/abstract",
• URL = "https://www.sciencedirect.com/science/article/pii/S1270963821005265",
• DOI = "doi:10.1016/j.ast.2021.107016",
• abstract = "This paper deals with the innovative development of surrogate models suitable for the simulation of aerodynamic performance and acoustic emission in terms of tonal components, of multi-propeller systems like those applicable in urban air-mobility vehicles. These can be of great help particularly when designing distributed-electric-propulsion configurations, as they provide an agile tool that avoids the need for computationally expensive CFD/CAA predictions. Without losing the generality of the conclusions that can be drawn about the capability of the proposed surrogate models to accurately describe multi-propeller aerodynamic and aeroacoustic responses, applications to a single propeller configuration are presented. Focusing on the simulation of the effects due to the spanwise distribution of blade twist and chord length, two surrogate modelling techniques are examined: one based on Artificial Neural Networks and one based on Genetic Programming. The numerical database for the identification of these models is determined by the combined application of a boundary integral formulation suitable for the potential aerodynamics solution around lifting/thrusting bodies, and the Farassat 1A formulation for the evaluation of the noise field. The numerical investigation demonstrates that both metamodelling techniques are able to reproduce propeller aerodynamic performance and radiated noise with a very good level of accuracy, certainly suitable for preliminary design applications",

}

Genetic Programming entries for Caterina Poggi Monica Rossetti Giovanni Bernardini Umberto Iemma Cristiano Andolfi Christian Milano Massimo Gennaretti

Citations