Increasing attention has recently been paid to the effects of localised pitting corrosion on the ultimate strength of marine structures. In this paper, an adaptive neuro-fuzzy inference system (ANFIS) method was developed to predict the ultimate strength reduction of steel plates with pitting corrosion subjected to uniaxial in-plane compressive loads. Published ultimate strength data-sets for unstiffened plates affected by pitting corrosion were used to train and test a series of ANFIS models composed of several input variables. To develop the best accurate model, rule-based fuzzy sets were used for mapping the inputs to the output using seven different types of membership functions. The two-sided Gaussian-type function was found to be more effective and less sensitive to the sample size than other functions tested. The developed method provided good estimates (maximum RMSE of 0.019) in comparison with published results obtained using the finite element and artificial neural network methods.
ANFIS method for ultimate strength prediction of unstiffened plates with pitting corrosion
Ships and Offshore Structures ; 13 , 5 ; 540-550
2018-07-04
11 pages
Article (Journal)
Electronic Resource
English
Ultimate strength of stiffened plates with pitting corrosion
Online Contents | 2015
|Simulation of the Ultimate Compressive Strength of Unstiffened Rectangular Plates
British Library Online Contents | 1993
|Ultimate Compressive Strength of Imperfect Unstiffened Plates: Simple Design Equations
Online Contents | 2004
|Computation of ultimate strength of locally corroded unstiffened plates under uniaxial compression
British Library Online Contents | 2007
|