<![CDATA[The planning of multi-story steel portal frame needs to watch for the resulted structure response due to the planning of earthquake-resistant building is needed in the earthquake-vulnerable area such as Indonesia. One of the method used to predict structure response of multi-story steel portal frame is Artificial Neural Network (ANN). The structure used to get the structure response is 10-story steel portal frame, modeled with the help of a element software and earthquake spectrum response analysis method according to SNI 1726-2012. Analyzing is conducted on each capital city of the 34 provinces with 3 different soil types, resulting in 102 data sets. It is therefore concluded that biggest values of movement response and structure velocity are, respectively, 0,0497 m and 0,0292 m/s in the city of Palu, and then the biggest value of structure acceleration is 2,15932 m/s2 on Palu. The accuracy level reaches 99% with 816 training data sets and Mean-Squared Errors (MSE) value is 0,00485. Therefore, it is concluded that ANN can predict multi-story steel portal frame response on all capital cities in Indonesia.Keyword : Multi-story steel portal frame, structure response, Artificial Neural Network]]>
Copyrights © 2017
