Optimization Of Cantilever Retaining Wall Design Using Improved Teaching-Learning-Based Optimization Algorithms

Abstract

Retaining structures play a crucial role in geotechnical engineering to support soil levels, prevent slope failure, and create flat surfaces for construction. Designing these structures involves optimizing internal and external stability while minimizing material usage and cost. This study focused on optimizing reinforced concrete cantilever retaining walls using the Teaching-Learning Based Optimization (TLBO) algorithm and an improved version (I-TLBO) with agents. In the context of the study, geometric-structural design variables, geotechnical -structural constraints, and optimization processes were examined. Minimizing weight and minimizing cost of the wall were the objectives considered in the cantilever retaining wall design process. The optimization results were compared with other algorithms in the literature, such as genetic algorithms, evolutionary strategies, and particle swarm optimization. The improved TLBO algorithm demonstrated superior performance, achieved lower design dimensions, and reduced costs. It provided more efficient solutions that pushed design constraints closer to their limits, resulting in a cost-effective and structurally sound cantilever retaining wall design. As a result of the study, the I-TLBO algorithm was found to be more cost and weight-effective than other methods in the optimization of cantilever retaining wall design.