Advances in Research

Abstract: An optimum design of non-linear steel frames using an effective artificial intelligence algorithm is presented. To consider the actual behavior of steel connections, the studied steel frames were designed as semi rigid connections. The Frye and Morris polynomial model is used for modeling the non-linear behavior of the semi-rigid connections. In this work, the Intelligent Tuned Harmony Search (ITHS) optimization algorithm was implemented due to its efficiency in parameter initializing through maintaining a proper balance between diversification and intensification throughout the search process. The design algorithm obtains the minimum weight of steel frames by choosing from a standard set of the AISC steel sections. Strength constraints of American Institute of Steel Construction - Load and Resistance Factor Design (AISC-LRFD) specification, deflection, displacement, size constraint and lateral torsional buckling are imposed on frames. To demonstrate the application and validity of the algorithm, this paper presents two steel frames with extended end plate without column stiffeners. The results reflect the superiority of the ITHS algorithm in terms of accuracy, convergence speed, and robustness when comparing with the state-of-the-art harmony search algorithm (HS) and Genetic algorithm (GA).