SOFT GROUND SUBSIDENCE PREDICTION OF HIGHWAY BASED ON THE BP NEURAL NETWORK

Abstract

Soft clay ground subsidence data of highway embankment Ipoh project in Malaysia use to build Back-Propagation artificial neural network model. The forecasts of soft ground subsidence final settlement find then comparing results of soft ground subsidence final settlement, then comparing the predict results with curve fitting hyperbola method, the curve method, three-point method forecast results. It turns out that neural network can avoid the human factors of interference from traditional methods, gaining high precision.