摘要：AbstractRoad measured data become useful information only if treated in order to match each section with statistically meaningful values. The problem can be led back to the detection of those change-points that divide measures in homogenous segments along the series of measured data. In this paper a methodology to detect a change point is proposed, searching those points that minimize the sum of the squared errors respect to the series of data. The MINSSE (MINimization Sum of Squared Error) has been, therefore, compared with the AASHTO Cumulative Difference Approach and a bayesian approach for the retrospective detection of change-points.
