APPLICATION OF NEURAL NETWORK MODELS FOR FORECASTING OF PAVEMENT CRACK INDEX AND PAVEMENT CONDITION RATING

Timely identification of undesirable crack, ride and rut conditions has been a critical step in pavement management at the network level. To date, many models have been developed for forecasting of pavement conditions with most of them focusing on a single index. Florida Department of Transportation (FDOT) forecasting models for roughness, skid resistance, or crack condition, etc., are such examples. The overall pavement surface condition is jointly determined by these individual pavement condition indices. This report summarizes the results of a research project that was initiated to implement a pavement condition prediction methodology using the Artificial Neural Network (ANN). In this research effort, three individual ANN models were developed to forecast three key indices, including crack rating, ride rating, and rut rating. These indices have been used by the FDOT for pavement evaluation purposes. Each individual model was trained and tested with the use of the FDOT pavement surface condition database. Modeling results suggest that the ANN models developed in the research have the capability to satisfactorily forecast future individual pavement condition index up to a period of five years. As one of the implementation tasks, a software package was developed for easy implementation and use of the ANN predictive models.

• Record URL:
  https://fdotwww.blob.core.windows.net/sitefinity/docs/default-source/research/reports/fdot-bc353-13rpt.pdf
• 
• Corporate Authors:

  University of South Florida, Tampa

  Department of Civil and Environmental Engineering, 4202 East Fowler Avenue
  Tampa, FL  United States  33620

  Florida Department of Transportation

  Haydon Burns Building, 605 Suwanee Street
  Tallahassee, FL  United States  32301
• Authors:
  • Yang, Jinglin
  • Lu, J J
  • Gunaratne, M
• Publication Date: 2003-3

Language

• English

Media Info

• Features: Appendices; Figures; References; Tables;
• Pagination: 152 p.

Subject/Index Terms

• TRT Terms: Condition surveys; Forecasting; Neural networks; Pavement cracking; Ride quality; Roughness; Ruts (Pavements); Skid resistance; Software packages
• Identifier Terms: Florida Department of Transportation
• Subject Areas: Design; Highways; Pavements; I22: Design of Pavements, Railways and Guideways; I23: Properties of Road Surfaces;

Filing Info

• Accession Number: 00942566
• Record Type: Publication
• Report/Paper Numbers: Final Report
• Contract Numbers: BC353-13
• Files: TRIS, ATRI, STATEDOT
• Created Date: May 19 2003 12:00AM