An improved constrained differential evolution for optimal design of steel frames with discrete variables

Abstract

New metaheuristic algorithms have been shown to be robust and effective for engineering optimization problems. Differential Evolution (DE), which is one of the most famous optimization approaches developed for optimal design problems with continuous design variables, is, in fact, inappropriate for constrained problems containing discrete design variables. In this article, an improved constrained differential evolution (iCDE) algorithm is proposed for optimization of steel frames with discrete design variables, followed by comprehensive testing of iCDE to validate the results. In order to make iCDE more efficient, two novel mutation approaches are proposed and tested. Three benchmark structural optimization problems are evaluated by the algorithm and compared against those reported in the literature obtained by GA, ACO, IACO, ICA, HS, and SBO. The total weight of the frame is considered as the objective function and the cross sections of the elements are taken as design variables. Constraints are based on the stress and displacement requirements conforming to the AISC-LRFD specifications. Moreover, to demonstrate the robustness of the algorithm, eight benchmark test functions of CEC-2017 are evaluated. The results show that the iCDE algorithm is computationally efficient and robust enough to find the global optima.

Keywords:

• Improved differential evolution
• discrete optimization
• metaheuristic algorithms
• steel frames
• structural optimization
• robust algorithm

Disclosure statement

The authors declare that they have no conflict of interest.