Determining hydraulic conductivity parameters of porous asphalt concrete using Bayesian parameter estimation

Abstract

We describe a technique to determine the hydraulic conductivity parameters of porous asphalt concrete using Bayesian parameter estimation. Three existing models were employed, and variables were assigned as statistical Bayesian parameters based on the Latin hypercube sampling approach. A thousand samples of randomly generated parameters were sorted, and samples of each variable were selected at stochastic intervals. A sequence of integers for each variable was generated to represent a random permutation of the integers, resulting in a set of sample parameters. Using these parameters, the three models were evaluated based on a comparison between the measured and predicted data. Finally, the appropriate model can be used to perform the stormwater management analysis of urban areas in terms of the probabilistic parameters obtained from the Bayesian updating scheme.