Displacement back analysis on supporting structure of deep foundation pit based on evolutionary neural nrtwork

Abstract

An evolutionary neural network method of displacement back analysis on supporting structure of deep foundation pit is proposed to search the optimal mechanical parameters. First, the BP network replaces the time-consuming finite element method to establish the non-linear relationship between the values of deep foundation pit mechanical parameters and displacement of its supporting structure, then genetic algorithm is used as an optimization method to search the optimal mechanical parameters in their global ranges. Application of this methodology is illustrated with a numerical example and reasonable results are yielded.