Proposing a novel oriented genetic algorithm for optimum seismic design of steel moment resisting frames

This paper proposes a new variant of the genetic algorithm (GA), called the oriented genetic algorithm (OGA), for optimum seismic design of steel special moment resisting frames. Since GA is mainly based on random operators, its computational burden is usually high. To overcome this issue, OGA takes advantage of the violation values of design constraints of the problem, such as elements’ demand-to-capacity ratios, strong column-weak beam requirements, and story drift ratios, to direct the search procedure. OGA is applied to a set of steel frames with different geometrical properties to demonstrate, and the results are compared to those of GA. The numerical results indicate that OGA significantly reduces the total number of function evaluations (NFE) required to obtain the optimum solutions. Also, the convergence history of OGA is compared with Particle Swarm Optimization and Ant Colony Optimization algorithms. It is shown that for a specified NFE limit, OGA gives better-optimized results.