TRB Pubsindex
Text Size:
x

« Prev Next »

Return to Results | New Search

Bookmark and Share

Title:

Performance Prediction of Interstate Flexible Pavement Across the Midwestern United States: Random-Parameter Regression vs Artificial Neural Network

Accession Number:

01698424

Record Type:

Component

Abstract:

Highway agencies in the Midwestern United States seeking to maintain their pavements in a state of good repair are facing increasing financial restraints. In this regard, the prediction of pavement performance plays a key role in efficient pavement management strategies. This paper compares performance predictions for Interstate flexible pavements from the Midwestern states (Indiana, Illinois, Wisconsin, Michigan, Ohio, Minnesota, Iowa and Missouri). Fixed and random parameter regression and Artificial Neural Network (ANN) models are estimated to predict the International Roughness Index (IRI), widely used as a performance indicator of pavements. Pavement performance data were obtained from the Long-Term Pavement Performance (LTPP) database of the Federal Highway Administration. Pavement age and freeze index were found to be statistically significant variables that influence the IRI. The ANN model, with a higher R², was found to outperform its fixed and random parameter regression counterparts. This was followed by a validation of models using out-of-sample data. A sensitivity analysis, using the trained ANN model, showed that an increase in pavement age and freeze index significantly affects pavement roughness. The model can assist highway agencies in carrying out performance-based scheduling of maintenance, repair, and rehabilitation (MRR) activity of the Interstate flexible pavements.

Supplemental Notes:

This paper was sponsored by TRB committee AFD10 Standing Committee on Pavement Management Systems.

Report/Paper Numbers:

19-02723

Language:

English

Corporate Authors:

Transportation Research Board

Authors:

Yamany, Mohamed S
Saeed, Tariq Usman
Volovski, Matthew

Pagination:

7p

Publication Date:

2019

Conference:

Transportation Research Board 98th Annual Meeting

Location: Washington DC, United States
Date: 2019-1-13 to 2019-1-17
Sponsors: Transportation Research Board

Media Type:

Digital/other

Features:

Figures; References; Tables

Identifier Terms:

International Roughness Index

Geographic Terms:

Midwestern States

Subject Areas:

Highways; Maintenance and Preservation; Pavements

Source Data:

Transportation Research Board Annual Meeting 2019 Paper #19-02723

Files:

TRIS, TRB, ATRI

Created Date:

Dec 7 2018 9:22AM