Predictive modeling of fatigue and rutting parameters for asphalt cement modified with pretreated oil palm clinker using artificial neural network algorithms to enhance pavement performance

Currently, the viscoelastic properties of standard asphalt cement cannot sustain the increasing demands resulting from heavier traffic loads, greater stress levels, and changing environmental conditions. Thus, the usage of modifiers is encouraged. Also, the Sustainable Development Goal (SDG) encourages the use of waste resources and emerging technologies in asphalt pavement technology. This study intends to harness this gap by examining the use of oil palm clinker (OPC) as an asphalt-cement modification to improve its viscoelastic properties using an innovative prediction approach. The modified asphalt-cement was produced by varying the acid-treated OPC powder (OPCP) content at 2%, 4%, 6%, and 8% and the rutting and fatigue performance was evaluated. This paper also presents an optimization approach and prediction comparison based on statistical modeling and artificial intelligence (AI) algorithms for the fatigue and rutting parameters of the modified asphalt cement. Model variables for the predictive models include OPCP content and test temperatures. The AI algorithms use 70% of the data for training, 15% for testing, and 15% for validation. The results showed that the incorporation of OPCP improves the properties of pure asphalt-cement by increasing stiffness and temperature susceptibility and that the crystalline phase of Si–O formed a novel structural group Si–OH. The RSM R 2 for rutting for unaged and RTFO aged responses was (99.743 and 99.893), the RMSE was (436.210 and 954.945), and the MRE was 3.269 and 2.315) for the model statistical performance index, respectively. The ANN R2 for rutting for unaged and RTFO aged responses were (99.903 and 99.970) the RMSE (106.283 and 528.500) and MRE (1.759 and 1.039). PAV fatigue RSM R2 values were (99.984), RMSE (77979.750), and MRE (12.089), while ANN R 2 values were (99.997), RMSE (53933.500), and MRE (5.262). The findings demonstrated that the generated model and algorithm could predict the fatigue and rutting performance of the OPCP-modified asphalt cement accurately with the AI algorithms model outperforming the statistical model. Also, the study aligns with SDG 9 by developing advanced modeling techniques and enhancing infrastructure durability through innovative use of modified materials as well as SDG 12 by incorporating recycled materials into sustainable production practices.