Modeling of asphalt pavement performance indices in different climate regions using soft computing techniques

Ali, Abdualmtalab Abdualaziz Yeklef (2022) Modeling of asphalt pavement performance indices in different climate regions using soft computing techniques. Doctoral (PhD) thesis, Memorial University of Newfoundland.

[img] [English] PDF - Accepted Version
Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission.

Download (13MB)


Pavement Management Systems (PMS) enhance pavement performance over the pavements' predicted lifespan by maximizing pavement life. PMS have become an essential aspect of construction and maintenance in the road domain, providing significant cost and energy emission reductions. In addition, using pavement performance prediction models have become an important part of PMS as a technically method for road engineers and various transportation agencies during the past several decades. The Pavement Condition Index (PCI) and International Roughness Index (IRI) are generally accepted methods for gauging ride quality and pavement conditions. Asphalt pavements are highly sensitive to various parameters, including pavement distress, environment, and traffic volume. Hence, studying these variables while developing prediction models is a vital step that can help develop asphalt pavement performance indices. This research aimed to introduce an effective method for developing asphalt pavement performance indices in different climate regions. This research provided a methodology to develop performance models using three soft computing techniques, namely the fuzzy inference system (FIS), multiple linear regression (MLR), and artificial neural networks (ANNs). Two sources were used for the extracted dataset: the long-term pavement performance (LTPP) data set for four climate regions in the U.S. and Canada and filed survey data of section roads of St. John's, Newfoundland, Canada. First, for the classification section, the research presented in this study provided a FIS that uses appropriate membership functions for computing PCI and IRI values. A fuzzy input was calculated by considering the degree of distress from nine density types of pavement distress coefficients (rutting, fatigue cracking, block cracking, longitudinal cracking, transverse cracking, patching, potholes, bleeding, and ravelling), which were considered as fuzzy input variables. Results presented that the rutting and transverse cracking had the most significant influence on the PCI model, while rutting and patching had the most significant impact on the IRI model. Second, the MLR and ANNs techniques were used for predicting and developing models for the PCI and IRI of flexible pavements. The LTPP database was used to obtain three fundamental variables (pavement distress, environmental, and traffic volume) as input variables for four climate regions. Finally, for the case study, the research developed a second set of pavement distress models based on a field survey of St. John's city's input variables for predicting PCI and IRI models. A high determination coefficient (R²), low root mean square error (RMSE) and mean absolute error (MAE)indicated good accuracy for the prediction models. The results showed that the ANNs have more precision than the MLR techniques. However, the results showed that both methods perform well.

Item Type: Thesis (Doctoral (PhD))
Item ID: 15883
Additional Information: Includes bibliographical references (pages 247-246)
Keywords: asphalt pavement performance, artificial neural networks (ANNs), Pavement Condition Index (PCI), International Roughness Index(IRI), Pavement Management Systems (PMS), fuzzy inference system (FIS)
Department(s): Engineering and Applied Science, Faculty of
Date: October 2022
Date Type: Submission
Digital Object Identifier (DOI):
Library of Congress Subject Heading: Pavements, Asphalt--United States; Pavements, Asphalt--Canada; Neural networks (Computer science); Pavements--Performance; Pavements--Maintenance and repair--Management

Actions (login required)

View Item View Item


Downloads per month over the past year

View more statistics