ANN based creep constitutive model for Marine sediment clay in coastal zone

Xiaobo Xiong,Liduo Zhou,Yan Chen,Ying Liu,Haifeng Dai
DOI: https://doi.org/10.1201/b15794-72
2013-01-01
Abstract:In the long coastal zone of Jiangsu province, there was widely distributed marine soft soil. As we all know, Marine sediment clay has higher creep characteristic. In this paper, according to the engineering practice information, systemic analysis of the engineering properties has been carried out. Based on the creep data of laboratory experiments, an empirical model was set up to fitting the rheological characteristic. Simultaneously, the principle and its application of neural network were introduced. Through improving the algorithm of Neural Network (NNT), the NNT constitutive model was built. It could avoid building a necessary complex constitutive equation in order to meeting the variable law of trial curves. The method was tested and verified by the creep experimental data of soft clay of Lianyungang. It is shown that the rheological model is rational and could effectively describe the nonlinear creep characteristic of marine sediment clay with better prediction. © 2013 Taylor & Francis Group, London.
What problem does this paper attempt to address?