In this paper, artificial neural networks (ANN)-based response surface method (RSM) is presented. This method is compared with conventional polynomial-based RSM in the context of structural reliability analysis. ANN-based RSM is then applied to predict ultimate strength of unstiffened plates. It is found out that the ANN-based RSM is more accurate and efficient than polynomial-based RSM in structural reliability analysis. ANN-based RSM can more accurately predict ultimate strength of unstiffened plates than the existing empirical formulae.
|Title of host publication||Proceedings of the Fifteenth (2005) International Offshore and Polar Engineering Conference, Vol 4|
|Publisher||The International Society Offshore and Polar Engineers|
|Publication status||Published - 2005|
|Name||International Offshore and Polar Engineering Conference Proceedings|
|Publisher||International Society Offshore & Polar Engineers|
- ANN-based response surface method
- artificial neural networks
- response surface method
- ultimate strength of plates
- design equations of plates
Pu, Y. C., Mesbahi, E., & El Hewy, A. H. (2005). ANN-based response surface method and its application to ultimate strength of plates. In Proceedings of the Fifteenth (2005) International Offshore and Polar Engineering Conference, Vol 4 (pp. 752-758). (International Offshore and Polar Engineering Conference Proceedings; Vol. 4). The International Society Offshore and Polar Engineers.