Development of an intelligent pavement analysis toolbox

Abstract

An intelligent pavement analysis toolbox was developed for backcalculating the mechanical properties of flexible pavements from non-destructive test data through an adaptive, multi-hybrid, integrated non-linear systems approach. The problem involves searching for the optimal combination of pavement layer stiffness solutions in an unsmooth, multi-modal, complex search space which makes it amenable to the application of differential evolution, a stochastic parallel direct search evolution strategy optimisation method. This innovative approach takes advantage of the combined efficiency and accuracy achieved by integrating advanced pavement numerical modelling schemes such as finite-element methods, computational intelligence-based surrogate mapping techniques such as artificial neural networks, and heuristics-based global optimisation strategies such as differential evolution, and yet provides a user-friendly automated pavement evaluation toolbox for the highway engineer to use on a real-time basis for accurate infrastructure evaluation. The main aim of the present study was to demonstrate the feasibility of using such an integrated modular systems approach to the complex non-linear parameter identification problem in pavement engineering.