Open AccessMulti-objective optimization of retaining wall using genetic algorithm
Author(s) -
Filip Dodigović,
Krešo Ivandić,
Jasmin Jug,
Krešimir Agnezović
Publication year - 2021
Publication title -
inženjerstvo okoliša
Language(s) - English
Resource type - Journals
eISSN - 1849-4714
pISSN - 1849-5079
DOI - 10.37023/ee.8.1-2.8
Subject(s) - embedment , genetic algorithm , bearing capacity , mathematical optimization , convergence (economics) , algorithm , optimization algorithm , computer science , structural engineering , engineering , mathematics , economics , economic growth
The paper investigates the possibility of applying the genetic algorithm NSGA-II to optimize a reinforced concrete retaining wall embedded in saturated silty sand. Multi-objective constrained optimization was performed to minimize the cost, while maximizing the overdesign factors (ODF) against sliding, overturning, and soil bearing resistance. For a given change in ground elevation of 5.0 m, the width of the foundation and the embedment depth were optimized. Comparing the algorithm's performance in the cases of two-objective and three objective optimizations showed that the number of objectives significantly affects its convergence rate. It was also found that the verification of the wall against the sliding yields a lower ODF value than verifications against overturning and soil bearing capacity. Because of that, it is possible to exclude them from the definition of optimization problem. The application of the NSGA-II algorithm has been demonstrated to be an effective tool for determining the set of optimal retaining wall designs.