Artificial Neural Network-Based Methodologies for Rational Assessment of211 Remaining Life of Exisiting Pavements. Development of a Comprehensive, Rational 211 Method for Determination of Remaining Life of an Existing Pavement

摘要

Most mechanistic-empirical methods for determining the remaining life of an211u001eexisting pavement rely on the use of deflection-based non-destructive evaluation 211u001e(NDE) devices. This report describes a methodology based on Artificial Neural 211u001eNetworks (ANN) techniques to estimate the remaining life of flexible pavements 211u001egiven the occurrence of two possible failure modes: rutting and fatigue cracking. 211u001eThe ANN techniques are also used to develop models that predict the critical 211u001estrains at the interfaces of the pavement. The inputs to all the models are the 211u001ebest estimates of the thickness of each layer and the surface deflections 211u001eobtained from a Falling Weight Deflectometer test. Uncertainty in these variables 211u001eis accounted for by the proposed methodology. The report also describes an 211u001eapproach to the production of pavement performance curves using the results of 211u001ethe ANN models.