Multi-Parameter Identification of Geomembrane Viscoelastic-Plastic Creep Constitutive Model by Genetic Algorithm

Abstract

Choosing an appropriate geosynthetics creep constitutive model is the key issue of determining its creep performance. In this paper, a creep overlay model with 7 parameters was put forward based on the analyses of current models. Moreover, the identification of this overlay model was realized by taking use of genetic algorithm MATLAB program. This method proved to be workable by practical example, so it provides a new idea for establishing geosynthetics constitutive model. Creep constitutive model has been the core issue in the research on geosynthetics performances. Although the model identification has gained rich achievements in theory, it is still a problem in practical application. In literature [1] the author adopted Laplace transform to prove that overlay model is strongly inclusive and various creep models could be obtained by selecting different parameter values. So the model identification could be transformed to parameters identification, and in fact both could be realized simultaneously. This avoids overdependence on specific model in the process of identification and the subjectivity of engineering calculation in choosing the constitutive model. This paper will combine overlay model and genetic algorithm to identify the geomembrane creep constitutive model.