INTERVAL ARTIFICIAL NEURAL NETWORK BASED RESPONSE OF UNCERTAIN SYSTEM SUBJECT TO EARTHQUAKE MOTIONS

Earthquakes are one of the most destructive natural phenomena which consist of rapid vibrations of rock near the earth’s surface. Because of their unpredictable occurrence and enormous capacity of destruction, they have brought fear to mankind since ancient times. Usually the earthquake acceleration is noted from the equipment in crisp or exact form. But in actual practice those data may not be obtained exactly at each time step, rather those may be with error. So those records at each time step are assumed here as intervals. Then using those interval acceleration data, the structural responses are found. The primary background for the present study is to model Interval Artificial Neural Network (IANN) and to compute structural response of a structural system by training the model for Indian earthquakes at Chamoli and Uttarkashi using interval ground motion data. The neural network is first trained here for real interval earthquake data. The trained IANN architecture is then used to simulate earthquakes by feeding various intensities and it is found that the predicted responses given by IANN model are good for practical purposes. The above may give an idea about the safety of the structural system in case of future earthquakes. Present paper demonstrates the procedure for simple case of a simple shear structure but the procedure may easily be generalized for higher storey structures as well.

Keywords:

earthquake , IANN , structure , building , Chamoli , Barkot.