سال انتشار: ۱۳۹۰
محل انتشار: ششمین کنگره ملی مهندسی عمران
تعداد صفحات: ۷
S. Gholizadeh – Assistant Prof. of Civil Engineering Department, Urmia University, Urmia, Iran
P. Torkzadeh – Assistant Prof. of Civil Engineering Department, University of Kerman, Kerman, Iran
S. Jabbarzadeh – M.S. Student of Kerman Graduate University of Technology, Kerman, Iran
In this paper, an efficient method is proposed for optimum shape design of double-layer grids. In optimization process, the weight of structure is considered as objective function. The design variables are the number of spans divisions of grid in two directions, the height of between two layers and the cross sectional area of elements. The design constraints are considered as limitations of the stress and slenderness of elements and the displacement requirements of joints. The optimization is carried out by particle swarm algorithm that is suitable for discrete and continuous variables. To reduce the computational time of optimization process, the structural responses are predicted using properly trained radial basis function neural network. This network is a robust network for predicting the structural responses. The numerical results demonstrate the robustness and high performance of the suggested method for the optimum shape design of double-layer grids in.