Applying Method of Moments to Model Reliability of Pavements Infrastructure

Abstract

Even though the concepts of pavement reliability and risk-based design have been recognized for a long time, probabilistic pavement design methods are often limited to the application of Monte Carlo simulation or Taylor expansion to evaluate reliability measures. This paper presents a comprehensive framework for establishing pavement reliability functions that consider various uncertainties associated with a pavement, starting from design assumptions to construction process and finally to utilization. With the recent development in the theory of structural reliability, in particular, the introduction of the standard normal point-estimate method for approximation of moments of the limit state function and the standardized functions for obtaining moment reliability indices, analytical evaluation of the probability integral for reliability has become feasible. The primary advantage of the analytical approach is its suitability for implementing optimization models of classical nonlinear algorithms. Finally, a numerical study was conducted to demonstrate the applicability of the method of moments for predicting pavement reliability.