Inverse analysis of in-situ geotechnical measurements using a genetic algorithm

Y. Malecot (1, 2) , S. Levasseur (3, 4) , M. Boulon (1) , E. Flavigny (1)

1 3SR - Laboratoire sols, solides, structures - risques [Grenoble]
2 RV - Risques, Vulnérabilité des structures et comportement mécanique des matériaux
3 Université de Liège
4 FNRS - Fonds National de la Recherche Scientifique [Bruxelles]

This paper is dedicated to the identification of constitutive parameters of the Mohr-Coulomb constitutive model from the horizontal displacements of a sheet pile wall. Two kinds of optimization algorithms are presented to minimize the error function, the first one based on a gradient method and the other based on a genetic algorithm. The efficiency of each algorithm related to the error function topology is discussed. Finally, it is shown that the use of genetic algorithms to identify the soil parameters seems particularly interesting when the topology of the error function is complex. © 2004 Taylor & Francis Group plc, London, UK. All rights reserved.

Mots clés: Algorithms; Errors; Genetic algorithms; Geomechanics; Gradient methods; Inverse problems; Numerical models; Optimization; Piles; Topology; Constitutive parameters; Error function; Horizontal displacements; Inverse analysis; Optimization algorithms; Sheet pile wall; Situ geotechnical measurements; Soil parameters; Parameter estimation

Domaines: Mécanique [physics.med-ph]

Y. Malecot, S. Levasseur, M. Boulon, E. Flavigny. Inverse analysis of in-situ geotechnical measurements using a genetic algorithm. NOMOG-9, Aug 2004, Ottawa, Canada. pp.223-228. ⟨hal-01998328⟩