Fatigue life prediction of asphalt concrete pavement using a harmony search algorithm

Abstract

This paper proposes an efficient Harmony Search (HS) algorithm for determining the material parameters of a fatigue cracking model based on Accelerated Pavement Testing (APT). APT is preferred over typical laboratory experiments and testing because APT can simulate field pavement performance. The main advantage of the HS heuristic algorithm is its ability to find the best global solution of a nonlinear complex function based on a stochastic random search instead of a gradient search that requires derivative information. Finally, this paper develops a fatigue cracking model according to the hypothesis of cumulative damage for the asphalt concrete layer and compares APT to existing models. In comparison with existing models, the APT model is found to have the capability of simulating the field performance because it considers the wandering effect, resulting in longer fatigue cracking life than those life computed by the models from existing lab experiments.