geneticalgorithm

Dierent passive dampers have been presented in order to improve structural performance, and various investigations have been undertaken to nd their optimum distribution by dierent methods. In this paper, optimal distribution of TADAS dampers to improve the performance of a 3-story steel moment frame is investigated. This frame has the same loading condition, mass, geometry, and beam and column material, as the moment resisting frame of a 3-story SAC building located in Los Angeles. The design of the members of the frame does not satisfy ASCE41-06 criteria. A nonlinear static procedure, according to ASCE41-06 instruction, is used to analyze the frame under seismic load.
Based on the concept of the uniform distribution of deformation (UDD) algorithm, in order to obtain optimal design, structural resistance elements should be transferred from strong to weak portions. In this study, a modied UDD algorithm is used to achieve optimal distribution of dampers. This algorithm acquires the optimal stiness distribution of TADAS dampers, with respect to the demand to capacity ratio (DCR) of the stories. The DCR of each story is equal to the maximum DCR of elements in that story, which is calculated based on ASCE41-06 regulations. First, the proposed algorithm assigns minimum stiness to all dampers, then, increases TADAS stiness in stories that have DCR greater than the allowable value, and vice versa. This process continues until uniform distribution for the stories DCR is achieved.
Genetic and PSO algorithms are types of heuristic algorithms. Optimal distribution of dampers using these algorithms is also obtained and their results compare with the UDD algorithm. Heuristic methods utilize a stochastic search to nd optimum solution. First, they generate a population of solutions, then, nd proper solutions, with regard to the objective function, and in the next steps, attempt to produce better population. Finally, they converge to the optimum design.
The results show that the optimum stiness distribution of TADAS dampers is obtained when the distribution of stories DCR becomes uniform. Also, the UDD alogorithm acquires optimum distribution of dampers in fewer numbers of analyses in comparison with heuristic methods, because this algorithm uses engineering intelligence instead of stochastic search to nd an optimum solution.