The paper explored the usefulness of Artificial neural network (ANN) in predicting the frame displacements under seismic load. The acceleration that is relatively easy to measure is used as the input value and the displacements that can be used to intuitively judge the condition of structures is used as the output value. The methodology utilized the universal function approximation ability of ANN for defining the relations between two data. For training of ANN, learning data consisting of acceleration and displacements are calculated from a verified finite element model under various seismic loads. The performance of the trained ANN was evaluated by comparing the displacements from ANN and FEM for seismic loads not used for training. The study showed that the ANN trained by various seismic loads can predicts the displacements from the acceleration for the new seismic loads. The trained ANN can be used for predicting the displacements of various buildings exposed to seismic loads in real time.
|Journal||IOP Conference Series: Materials Science and Engineering|
|Publication status||Published - 2018 Nov 15|
|Event||14th International Conference on Concrete Engineering and Technology, CONCET 2018 - Kuala Lumpur, Malaysia|
Duration: 2018 Aug 8 → 2018 Aug 9
Bibliographical noteFunding Information:
This research was supported by the EDucation-research Integration through Simulation On the Net (EDISON) Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (NRF-2014M3C1A6038855).
All Science Journal Classification (ASJC) codes
- Materials Science(all)