Cost-effective and performance-optimized reinforced concrete retaining walls through differential evolution algorithm

Abstract

This study investigates the optimal design of counterfort retaining walls through the application of a Differential Evolution (DE) algorithm. A typical counterfort retaining wall comprises four fundamental components: stem, toe, heel, and counterfort. By treating the dimensions of these elements and the associated reinforcements as design variables, the optimal design process identifies the most cost-effective dimensions while adhering to the various constraints. The DE algorithm, a population-based optimization technique similar to Genetic Algorithms, distinguishes itself through its unique methodologies for crossover, mutation, and population updating. The construction cost of a retaining wall primarily encompasses the expenses for concrete, reinforcement steel, and formwork. In this study, the wall geometry was optimized using the DE algorithm, with the optimization framework implemented in MATLAB software. The computed results were compared with the recommended values for different wall heights. To ascertain the optimal combination of feasible design variables, objective functions were employed, contingent on the design variable values. This investigation utilized 12 design variables and 12 design constraints to optimize the objective function. Counterforts are incorporated to enhance the stability of the main wall, with a minimum thickness defined to ensure compliance with the specified lower limit values. Furthermore, the objective function was formulated for wall heights of 6, 7, 8, 9, and 10 m above ground level using the DE algorithm. The results demonstrate that the optimization of counterfort retaining walls can significantly reduce construction costs.