DEVELOPMENT OF POS PREDICTION MODELS USING PROJECT, HISTORICAL AND PROJECT-GENERATED QUALITY ASSURANCE DATA

The KDOT Pavement Management System includes a Project Optimization System (POS). For each pavement distress, the POS requires models which predict the probability that a given distress level occurs at some time in the future. An interim regression model was developed during the initial development of the POS. These models were considered interim subject to modification and/or replacement as better mechanistic models, statistical models, or historical data becomes available. This K-TRAN project investigated the mechanistic and empirical-mechanistic models and techniques advanced in research conducted by the National Cooperative Highway Research Program (NCHRP) and the Strategic Highway Research Program (SHRP). In addition, the project specific data developed for the models were reanalyzed using SHRP sensitivity analysis and regression procedures to determine if improved models could be developed. The statistical studies were successful in developing alternative regression models. However, the binary nature of the response variable imposes a significant limitation on the ability of regression techniques to estimate probabilities of accurance between these two limits.

Language

  • English

Media Info

  • Features: Appendices; Figures;
  • Pagination: 100 p.

Subject/Index Terms

Filing Info

  • Accession Number: 00732042
  • Record Type: Publication
  • Report/Paper Numbers: K-TRAN:KU-94-2
  • Files: TRIS, ATRI, STATEDOT
  • Created Date: Feb 27 1997 12:00AM