Optimal seismic design of reinforced concrete moment-resisting frames using an improved metaheuristic and neural networks

Performance-based seismic design optimization of reinforced concrete (RC) frames is a complex and computationally intensive problem in structural engineering. In this paper, a hybrid combination of soft computing techniques is proposed to efficiently deal with the seismic design optimization of RC frames. A new Chaotic Center of Mass Optimization (CCMO) algorithm is introduced to efficiently explore the design space. Additionally, an efficient neural network (NN) model is utilized to predict nonlinear structural seismic responses during the optimization process. The proposed hybrid methodology is applied to the minimum cost design of 5and 10-storey RC moment-resisting frames in the context of seismic performance-based design. The obtained numerical results indicate that the proposed methodology is a powerful tool for the seismic optimization of RC moment-resisting frames spending a reasonable computational cost.