Prediction the Dynamic Modulus of Hot Asphalt Mix Using Genetic Algorithms and Neural Network Modeling
Shadi Hanandeh, Aneke Frank Ikechukwu, Zaid Alajlan, Shadi Al Khateeb, Ruba A. Alkharabsheh
Abstract
The dynamic modulus is a fundamental characteristic of asphalt concrete and expresses the stiffness properties of a hot mix asphalt mixture as a function of temperature and loading rate. This study used artificial neural network modeling and genetic algorithms to evaluate the asphalt concrete dynamic modulus. The experimental database was collected from LTPP DATA that used in the ANN and genetic algorithm development and modeling. The output for the two models was the asphalt concrete dynamic modulus. Moreover, mathematical models were employed to predict the dynamic modulus of asphalt concrete with different parameters. Following the establishment of the model designs, the deficiencies and strengths of the proposed models are evaluated using determination coefficient (R2) values. The evaluation was performed by comparing the dynamic modulus of asphalt concrete predicted from four models with the dynamic modulus obtained from the experimental testing. Notably, the neural network models achieved precise calculations for models 1 and 2, with R2 values of 0.96 and 0.93, respectively. The genetic algorithm models achieved R2 values of 0.73 for model 1 and 0.64 for model 2. The two models, the genetic algorithm model and the artificial neural network model, contributed to the generation of two new empirical equations.