Skip to content

Important notice

The building where our contact centre is based was evacuated last week. Our 0800 numbers are up and running again, but please be patient as we have limited support available for our contact centre staff. We’ll do our best to help you if you need to call us.  For more information read our latest media release.

Access keys for nzta.govt.nz

  • h Home
  • m Menu
  • 0 Show list of access keys
  • 2 Skip to content
  • 3 Skip to top
Back to Resources

Research Report 212 Evaluating sensitivity of parameters in predictive pavement deterioration modelling

Published: | Category: Activity management , Research programme , Research & reports | Audience: General

This report describes the results of a project to investigate sensitivity of the input parameters used for NZ dTIMS system.

The objective was to identify the sensitive parameters affecting the output of the dTIMS analysis which will help the user on deciding which data item should be given higher priority in terms of accuracy level with which data are to be acquired.

It was also envisaged that the study will establish which of the parameters are effectively inactive in the pavement deterioration and maintenance programme generation process.

Five RAMM databases from different road controlling authorities were used to define the range of values for the data item used in NZ dTIMS system. The tradition ceteris paribus method (considering sensitivity of one parameter without inter-reaction with other parameter) and the Factorial Latic Hypercube method (considering the sensitivity of input parameters with inter-relation with other parameters) were used for the analysis.

The analysis was done in three stages:

  1. pavement deterioration prediction
  2. strategy generation
  3. economic optimisation.

Non sensitive parameters related to each stage were eliminated from consideration in the next stage.

The results showed that not all the parameters are sensitive to the modelling and only a few are quite sensitive. A list of the sensitive parameters with the magnitude of the sensitivity together with the recommendation on the data acquisition method was prepared.

Keywords: condition rating, data sensitivity, treatment selection, roads, roughness, statistics, predictive modelling, economic optimisation

Publication details

  • Author:
  • Published: 2001
  • Reference: 212
Top