Applying an Artificial Neural network- Developed Collective Animal Behavior Algorithm for seismic reliability evaluation of structure

Abstract

The seismic design contains considerable uncertainties, which originate from the structural geometries, earthquake motions, analytical models, and material properties. By considering all main uncertainties, reliability analysis is applied estimating the possibility of failure in each of a set of performance requirements. Structural failure localization, quantification, and failure possibility evaluation are principal outputs in the structures reliability evaluation. In this paper, an Artificial Neural Network- Developed Collective Animal Behavior Algorithm is used for reducing the problem complexity needed for reliability investigation and failure detect. In this way, a definite structure is modeled and some failure case are determined. These cases are recognized as train databases for organizing the ANN-DCABA. Therefore, the correlation among structure's response as input and structure's stiffness as output is provided utilizing Artificial Neural Network- Developed Collective Animal Behavior Algorithm. The provided method is so efficient and produces advisable precision in structural failure discovery under a series of ground motions. Besides, for assessing the structure reliability, 5 accidental variables are determined. These are columns’ state of the 1st, 2nd, and 3rd floor, gravity loads, and elasticity modulus. Finally, the Artificial Neural Network- Developed Collective Animal Behavior Algorithm specifies the failure probability of the proposed structure and it can reduce the computational effort required for reliability analysis and damage detection. However, MCS could indicate the failure probability for a given structure; the Artificial Neural Network- Developed Collective Animal Behavior Algorithm helps simulation techniques for receiving an acceptable precision and decrease computational effort.