Among the materials, which are used for paving roads, asphalt concrete plays an essential role.
In the following, we try to investigate the reasons.
Introduction of Asphalt Concrete
A key pavement design criterion in the Mechanistic-Empirical Pavement Design Guide is the international roughness index (IRI) for asphalt concrete pavement accuracy layers.
Studies have revealed that the MEPDG's IRI transfer function is just a linear combination of road parameters, making it unable to make precise predictions.
This study created an AdaBoost regression (ABR) model to enhance IRI's predictive power and compared it to the linear regression (LR) in MEPDG to address the problem.
The 4265 records from the Long-Term Pavement Performance (LTPP), which are trustworthy data that have been retained over years of monitoring, were used to create the ABR model.
Different kinds of Asphalt Concrete
Due to the effect of the construction environment, the longitudinal profile of the pavement frequently varies during the building of asphalt concrete (AC) roads.
In addition, over time and in difficult conditions, the outline shape will continue to change while driving, becoming visually apparent as the pavement gets rougher.
In addition to impacting the comfort and safety of the passengers and drivers, vehicles operating on uneven roads also increase operating costs (such as fuel consumption, reduced driving speed, and extended travel time)
And, concurrently, will hasten the deterioration of the pavement structure, impacting the service life and maintenance cycle of the pavement.
Specifications of Asphalt Concrete
This is not a novel idea that has not been investigated by several studies and analyses by relevant scholars from around the world.
The artificial intelligence (AI) approach can include nonlinear computations that are appropriate for creating models for predicting AC road distress.
For instance, Lin et al. discovered a three-layer ANN model in 2003 utilizing the data gathered for deep learning, with 14, 6, and 1 neuron in each of the input, hidden, and output layers, respectively, and examined the association between variables.
Chandra et al. created three models to investigate if the ANN model outperformed a single linear or nonlinear regression model.
The Price of Asphalt Concrete Pavement
The authors of this study evaluated and assessed an ABR model for IRI estimate in AC pavements with the goal of maximizing the accuracy and reliability of the estimation while also maximizing the model's design.
Weak learners, loss functions, and other factors of modifying the ABR's accuracy have all been taken into account in order to develop recommendations for effective model optimization.
The considered model data comes from the LTPP large-scale pavement information database.
It contains several elements, such as structure, climate, traffic, and performance variables that have an impact on the performance of the pavement structure.
Things you should know about Asphalt Concrete
Additionally, the outcomes demonstrate that the IRI0 (initial IRI) is the most significant factor, which is compatible with the MEPDG transfer function and enables a practical explanation for the ABR model.
The findings and analysis are beneficial for refining the IRI prediction model's performance evaluation of the pavement design structure.
It is advised to carry out the additional study to enlarge the data set, gather more precise road performance data and maximize the model's potential.