Stochastic Analysis of Laterally Loaded Pile using Artificial Neural Networks

Abstract eng:
Reliability analyses of a laterally loaded pile are carried out by coupling probabilistic methods with a standalone finite-element analysis package via an alternative approach based on the artificial neural network (ANN). The proposed approach consists of two steps. The first step obtains the design point and the reliability index using the response surface approach based on the linear polynomial function. The second step constructs an approximation of the performance function using the neural network approach based on the design point found in the first step. The proposed approach is illustrated for the case of a laterally loaded pile involving correlated non-normal basic random variables and accounting for one-dimensional spatial autocorrelation of soil resistance. Stochastic analyses are carried out on pile-head deflection failure and bending moment failure. The probabilities of failure are obtained using first-order and second-order reliability methods, and Monte Carlo simulations based on importance sampling.

• Export as: BibTeX | MARC | MARCXML | DC | EndNote | NLM | RefWorks
• Add to your basket