Estimation of the rutting performance of Polyethylene Terephthalate modified asphalt mixtures by adaptive neuro-fuzzy methodology

• Effect of PET modification was assessed on rutting properties of asphalt mixture.
• Environmental conditions were simulated.
• Dynamic creep test was performed.
• Adaptive neuro-fuzzy inference system (ANFIS) was selected as alternative to analytical approach.

In this paper, the accuracy of soft computing techniques was employed for prediction of the rutting performance of Polyethylene Terephthalate (PET) modified asphalt mixture. The process, which simulates the mixture’s deformation, was constructed with adaptive neuro-fuzzy inference system (ANFIS). The inputs were PET percentages, stress levels and temperatures. The performance of proposed system was confirmed by simulation results. The ANFIS results and the results achieved by experiments were compared using root-mean-square error (RMSE) and coefficient of determination. The experimental outcomes suggested that ANFIS approach can be used to improve predictive accuracy and capability of generalization.