Optimum design of real-size reinforced concrete bridge via charged system search algorithm trained by Nelder-Mead simplex

Abstract

In this paper, an optimization method based on the combination of Nelder-Mead (NM) simplex concepts and charged system search algorithm (CSS) is proposed and called NM-CSS. NM, as a search technique, involves the iterative formation and transformation of a geometrically arranged set of points known as a simplex to uncover the optimal solution. By incorporating NM into metaheuristic algorithms, the convergence speed and solution quality can be notably boosted. Throughout each iteration, the solutions undergo various operations, including reflection, contraction, and expansion, to enhance the effectiveness of the algorithm. In this algorithm, produced solutions by the CSS are sorted and divided into two groups of good and bad solutions, and then the NM operators are applied to bad solutions to generate the possible best ones. The proposed algorithm is verified through some mathematical benchmark functions. Then, it is implemented on a real-size reinforced concrete bridge to reach an optimum design. To recognize the effective parameters in the design of structural components of reinforced concrete bridges, a sensitivity analysis is carried out. The total cost of materials in piers and the deck is defined as an objective function, and the cross-section area of structural elements and longitudinal reinforcements are chosen as design variables. The results of simulations represent the stability and robustness of the proposed NM-CSS method compared to standard CSS. In other words, utilizing the proposed NM-CSS algorithm in an optimum design of the piers and deck results in saving %3.685 and %2.084 of costs, respectively, in comparison with the results of the standard CSS.