Integrating Rolling Wheel Deflectometer Deflection Measurements into Pavement Management Systems using Multivariate Statistical Methods and Fuzzy Logic

摘要

The Louisiana Department of Transportation and Development (LADOTD) conducted a researchproject on approximately 1400 miles of asphaltic concrete roadways with the Rolling WheelDeflectometer (RWD). The major outcomes of the study were to assess the effectiveness of theRWD as a structural assessment tool on Louisiana Highways and to introduce the RWD index(RI).The purpose of this paper was to define RI ranges for the treatment selections orpavement structural conditions linked to current LADOTD PMS methods using multivariate(Factor Analysis), descriptive (Analysis of Variance (ANOVA)) statistical methods, and fuzzylogic-functions.The Factor Analysis revealed that RI could be generally used as a single index to assessthe pavement structural conditions. The ANOVA analysis revealed that RI could not be used toestablish the four treatment methods typically used by LADOTD. However, the ANOVAanalysis did reveal that the RI could be used to infer whether a pavement was structurally-soundor deficient. If determined to be structurally- deficient, additional structural evaluation would berequired such as with an FWD.Fuzzy logic/set theory was utilized to establish fuzzy functions for the structurally-soundand structurally-deficient pavement conditions along with RI threshold ranges per thicknessgroup. Equations were derived to produce RI threshold values for pavement thicknesses lessthan 12 in. The equations may be used for pavements thicker than 12 in., but with caution. Forlarge segments of roadways, algorithms were provided to assess whether pavements werestructurally- sound or structurally-deficient based on RI.