An inverse method of load determination using hybrid genetic algorithms, artificial neural networks and finite element analysis

This source preferred by Siamak Noroozi and John Vinney

Authors: Rabbani, M., Noroozi, S., Vinney, J. and Shrazi Kia, S.

Editors: Topping, B.H.V., Montero, G. and Montenegro, R.

Publisher: Civil-Comp Press

Place of Publication: Kippen

This data was imported from Scopus:

Authors: Rabbani, M., Noroozi, S., Vinney, J. and Shirazi Kia, S.

ISBN: 9781905088096

Load determination from structural response is considered as an inverse problem in structural engineering. In this paper, a hybrid genetic algorithm, neural network and finite element method for load identification on a component in service is proposed and described. The problem of training neural network is formulated as an optimization problem, which is then solved by using genetic algorithms (GA). Using genetic algorithms avoids some of the weaknesses of traditional gradient based search methods. Finite Element Method is used for generating data needed for training the hybrid system. The system is trained using strain data as input and known applied load as output. Once the hybrid system is trained, the estimated load can be obtained quickly and accurately. The numerical experiments suggest that good prediction of the load value and its location are possible and the proposed method is feasible. © 2006 Civil-Comp Press.

The data on this page was last updated at 05:17 on May 25, 2020.