Development of International Roughness Index from Pavement Distress Using Artificial Neural Networks (ANNs)

Ali, Abdualmtalab Abdualaziz; Esekbi, Mohamed Imbarek; Sreh, Muftah Mohamed

doi:10.21608/ijaebs.2023.165192.1054

Development of International Roughness Index from Pavement Distress Using Artificial Neural Networks (ANNs)

Document Type : Original Article

Authors

¹ Lecturer, Azzaytuna University, Civil Engineering Departement, Tarhuna, Libya

² Professor, Tripoli University, Civil Engineering Departement, Tripoli, Libya

³ Lecturer, Elmergib University, Civil Engineering Departement, Al-Khoms, Libya

10.21608/ijaebs.2023.165192.1054

Abstract

Roughness index forecasts are essential for optimizing pavement rehabilitation and treatment programs. The main objective of this study is to investigate the effect of pavement distress on pavement performance and develop International Roughness Index models (IRI) for dry no freeze regions in the U.S. Data for this research was collected from the Long-Term Pavement Performance (LTPP) database. The data include a total of 138 records of pavement distress with no maintenance and rehabilitation. Based on these data, IRI prediction models were developed using two modelling approaches: Multiple Linear Regression analysis (MLR) and Artificial Neural Networks (ANNs). The proposed models predict the IRI as a function of pavement distress variables such as or including fatigue cracking block cracking, edge cracking, longitudinal cracking, transverse cracking, potholes, patching, bleeding, and ravelling. This study showed that the (ANNs) model yielded a higher prediction accuracy than the (MLR) model.This study showed that the (ANNs) model yielded a higher prediction accuracy than the (MLR) model.

Keywords