|
چکیده
|
Record selection plays crucial role in the nonlinear analysis of structures. Several approaches have been emerged in this area of research in which the Generalized Conditional Intensity Measure (GCIM) is one of them. This approach takes the (log-normal) distribution of different (scalar) Intensity Measure (IM) into consideration. Then the records which have the closest match with these IM distributions are selected. This approach is comprehensively assessed in this paper in order to examine its efficiency. A set of 267 pairs records are selected with the magnitude range greater than 5 and the distance less than 100 km. Seventeen IMs are taken into account and their (log-normal) cumulative distribution functions are calculated. The considered IMS are: Spectral acceleration at T= 0.1, 0.2, 0.3, 0.5, 1.0, 2.0, 5.0, 10.0, Peak Ground Acceleration (PGA), Peak Ground Velocity (PGV), Arias Intensity(IA), Cumulative Absolute Velocity (CAV), Displacement Spectrum Intensity (DSI), Acceleration Spectrum Intensity (ASI), Spectrum Intensity (SI) and Duration (Ds575, Ds595). Then, eight records are selected among the whole dataset in which their CDFs have the minimum deviations when compared with the CDFs based on the whole dataset. The Genetic Algorithm (GA) is employed in order to solve this optimization problem. The response of the given structure was calculated by means of the incremental dynamic analysis. The results show that although the selected records have fully compatible CDFs with the CDF based on the whole records, however, the structural response is meaningfully different. This difference is more significant in the 16th and 84th fractiles.
|