In the conventional method of designing structures, a number of constants often characterize decisive parameters such as dimensions, material strengths and loads. Insufficient precision in modeling and implementing the structures are among the uncertainties that exist in structural design. If the maximum safety is expected for a structure, a combination of optimization and reliability should be used; this means that the structure should be optimized on the basis of reliability rather than economic considerations. In this study, geometry optimization of different steel frames has been done based on reliability; in fact, the objective is to obtain dimensions which are expected to have the best performance in terms of safety. To do so, different failure surfaces should firstly be defined and the structure should be optimized on the basis of the maximum reliability possible. The answer to the problem is the maximum distance from the defined failure surface obtained by probabilistic method. Also, decision-making analysis, which is the principle of engineering judgment, has been defined in the mathematical model so that the structural engineer has more options to determine the type of structural performance. The solution of linear and non-linear problems can be determined by implementing the algorithm for several times. Genetic algorithm can obtain the design point and probability of failure in statistical problems; therefore, it has been used to find the best dimensions of the frame under study.
کلید واژگان :reliability, optimization, decision making, genetic algorithms, opensees
ارزش ریالی : 600000 ریال
با پرداخت الکترونیک