Two-stage multi-objective optimization of reinforced concrete buildings based on non-dominated sorting genetic algorithm (NSGA-III)

Abstract

The colossal environmental footprint of the construction industry is a paramount ecological concern, especially in the wake of climate change threats and natural resources depletion. The present study performs a novel two-stage optimization approach for the design of reinforced concrete (RC) frame buildings to minimize the water footprint (WF) and cost, based on non-dominated sorting genetic algorithm (NSGA-III). The first stage includes multi-objective optimization (MOO) with respect to the cost and capacity of cross-sections of RC beam and column elements for any possible combination of design variables using section-based database schema (SBDBS). The second stage is MOO based on NSGA-III to minimize the volumes of reinforcing steel and concrete in RC structure considering the optimal sections generated by SBDBS as a pre-determined list for elements. Four structural archetypes consisting of 4, 8, 12, and 16-story frames were considered in the design optimization problems. The main objective of this optimization is to calculate the cost and WF of RC structures for various unit cost and WF of materials, and as a result, identify the optimal design of RC structures. The result of this optimization framework allows the decision maker to identify the optimal RC structural design for any cost or WF condition by performing a sensitivity analysis, without need for additional structural analysis or optimization. The proposed approach further allows the decision maker to consider the uncertainty of WF and cost of materials that emanate from regional or spatiotemporal variables. The results show that the shape of the optimal design Pareto front (PF) is more related to the values of unit WF of materials than the costs. To highlight the advantages of the proposed procedure, probabilistic sensitivity analysis and decision-making strategy were performed and pertinent conclusions were drawn.