Selection and scaling of ground motion records using harmony search

Abstract

In this study, a solution algorithm which is based on meta-heuristic harmony search algorithm is used for selecting and scaling ground motion records for the use in time history analysis. By means of the applied algorithm, ground motion record sets are obtained by selecting and time-domain scaling of real records complying with the elastic design spectrums defined in The Turkish Earthquake Code for different soil types. Firstly, the traditional approach is utilized and ground motion record sets are generated by selecting 199 ground motion records from PEER (Pacific Earthquake Engineering Center) Strong Motion Database considering the magnitude, distance and ground conditions. Selecting and scaling of ground motion record sets are performed by using these records. The results indicate that, various ground motion record sets can be obtained which are compatible with the requisite constraints by using harmony search algorithm. In conclusion, it is seen that harmony search algorithm can be used as an effective method for selecting and scaling real ground motions.