Detecting homogeneous pavement section using econometric test for structural changes in linear model

Pavement Management Systems require a preliminary evaluation of road condition data in order to detect homogeneous sections. If surveyed data are regarded as "time series”, they may be adequately described by linear dynamic processes with structural changes; where the change of the coefficients in the model shows the break between two homogeneous sections. The problem of detecting structural changes in linear regression relationships has been an important topic in statistical and econometric research. The most important classes of tests on structural change are those derived from the generalized fluctuation test framework, on one hand, and tests based on F statistics, on the other. In this paper, methodological issues related to the application of above mentioned econometric criteria to detect homogeneous road sections, are examined. The effectiveness of the econometric criteria were verified through the analysis of a case study related to skid resistance measurements within which a comparison between the different approaches has been performed. The implementation of these methods in the sectioning process shows that a reasonable breakdown of the road using new econometric principles is possible.