Gray Clustering-Based Pavement Performance Evaluation

In the Mechanistic Empirical Pavement Design Guide (MEPDG), rather than producing a quantitative overall evaluation index, several performance measures are used to separately evaluate the pavement performance condition for rehabilitation design. Due to the characteristics of information on parameters being partially known during the evaluation process, there exist several limitations of the current evaluation methods. In this paper a gray clustering-based methodology is proposed to comprehensively evaluate existing pavements following the MEPDG framework. Gray clustering method is one pillar part of the gray system theory, which was proposed in the 1980s to solve problems with partially known information. The performance criteria used in this study are selected from those defined in MEPDG. The analysis data sets are obtained from the long-term pavement performance (LTPP) database for three types of pavements: flexible pavement, jointed plain concrete pavement, and continuous reinforced concrete pavement. Many LTPP pavement sections are evaluated using the proposed methodology and comparisons are made between gray clustering-based method and extension theory-based approach. It is shown that the proposed gray clustering-based method is able to quantify pavement performance and is suitable to be used as a practical pavement evaluation tool for highway agencies.