Section based probabilistic performance prediction: fiction or future?

Life cycle cost analysis is a common method to assess maintenance strategies over a given period of time. Therefore performance prediction models are essential to evaluate the future condition of different attributes (rutting, cracking, etc.). Most pavement management systems use deterministic performance prediction models, which describe the future condition by a functional correlation between the condition attribute (technical parameter or index) and the descriptive variables (age, traffic load, temperature, etc.). For the practical application these models need detailed information about its variables but do not guarantee the quality of the prediction. To include the distribution of the descriptive variables and to get a conclusion about the quality of the performance prediction the model development must be based on a more probabilistic approach. In the context of a Masters thesis, which was carried out at the Vienna University of Technology in 2004, section based probabilistic performance prediction models using Markov-chains were tested in practice in a small community in Austria for the first time. The models were applied on different assets (carriageway, sidewalk, curbs, etc.). An overview is provided of the advantages and disadvantages, mathematical problems, and how section based probabilistic performance prediction models can be applied in practice on other road networks and assets in the future. For the covering abstract see ITRD E144473.

  • Authors:
    • WENINGER-VYCUDIL, A
    • SAMEK, G
    • ROHRINGER, T
  • Publication Date: 2008

Language

  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 01323367
  • Record Type: Publication
  • Source Agency: TRL
  • Files: ITRD
  • Created Date: Dec 22 2010 8:48AM