Reinforced concrete frame optimization using an adaptive particle swarm algorithm

This paper proposes a methodology for the optimization of reinforced concrete frames with moderate energy dissipation capacity using adaptive particle swarm optimization. The methodology proposed herein tackles the optimization problem formulated as: I) objective function is the frame cost; II) optimization variables, i.e. the dimensions and reinforcements of beams and columns; and, III) constraints stemming from the Colombian structural design code. To assess the proposed methodology, three planar frames were used, and the algorithm was run thirty times for each frame. The algorithm produced feasible designs for all three test frames, obtaining cost reductions between 30 to 70% (compared to the best initial solution cost). Another significant observation relates to the number of optimization variables: as the number of elements in the structure increases, so too does search space complexity. This increment is observed in the number of structural configurations that present a similar cost, which points to the existence of several local minimums.