AUTOMATED DETERMINATION OF CONSTITUTIVE PARAMETERS USING A METAHEURISTIC OPTIMIZATION METHOD IN SYNTHETIC EXPERIMENT

AUTOMATED DETERMINATION OF CONSTITUTIVE PARAMETERS USING A METAHEURISTIC OPTIMIZATION METHOD IN SYNTHETIC EXPERIMENT

Paper in proceedings (conference paper)

Higher complexity of modern constitutive models often results in an increase of number of input parameters and time requirements for their calibration. Optimization methods provide a possible solution for this problem. A program application based on the metaheuristic optimization method Particle swarm optimization (PSO) is presented in the paper. The application consists of two parts: a subroutine with the PSO algorithm and a subroutine involving a finite element (FE) model of an analyzed construction. The subroutine with PSO governs input parameters modifications and evaluation of the objective function. Modified values of input parameters are then used in the FE model subroutine for calculation. Different alternatives of the PSO method are reviewed and tested in the first part of the paper. In the second part, a synthetic numerical experiment of pressuremeter test is used for automatically driven determination of input parameters for two constitutive models: Mohr – Coulomb, Hardening soil. The procedure is able to perform fully automatic determination of input parameters values. However, when only single test is used, several combinations of input parameters might get to the same result. Future research will be therefore focused on a combination of multiple measurements (e. g. from different depths) used during the optimization task.

Higher complexity of modern constitutive models often results in an increase of number of input parameters and time requirements for their calibration. Optimization methods provide a possible solution for this problem. A program application based on the metaheuristic optimization method Particle swarm optimization (PSO) is presented in the paper. The application consists of two parts: a subroutine with the PSO algorithm and a subroutine involving a finite element (FE) model of an analyzed construction. The subroutine with PSO governs input parameters modifications and evaluation of the objective function. Modified values of input parameters are then used in the FE model subroutine for calculation. Different alternatives of the PSO method are reviewed and tested in the first part of the paper. In the second part, a synthetic numerical experiment of pressuremeter test is used for automatically driven determination of input parameters for two constitutive models: Mohr – Coulomb, Hardening soil. The procedure is able to perform fully automatic determination of input parameters values. However, when only single test is used, several combinations of input parameters might get to the same result. Future research will be therefore focused on a combination of multiple measurements (e. g. from different depths) used during the optimization task.

constitutive models, finite element method, pressuremeter test, metaheuristic, Particle swarm optimization

constitutive models, finite element method, pressuremeter test, metaheuristic, Particle swarm optimization

KOUDELA, P.; CHALMOVSKÝ, J.

2019

02.07.2018

Albena, Bulgaria

978-619-7408-36-2

18th International multidisciplinary scientific geoconference SGEM 2018 - Conference proceedings VOLUME 18 - Science and Technologies in Geology, Exploration and Mining Issue: 1.2 - HYDROGEOLOGY, ENGINEERING GEOLOGY AND GEOTECHNICS

1314-2704

International multidisciplinary geoconference SGEM

18

2

Republic of Bulgaria

121

128

8