Adaptive logic applications in pavement performance modeling

摘要

Pavement performance is one of the most important components of the pavement management system. Models for predicting pavement performance have been developed on the basis of traffic and time-related models, interactive time traffic or distress models. The characteristic feature of the models is that they are formulated and estimated statistically, recently however artificial neural networks are being used. The purpose of this paper is to extend the use of the adaptive logic networks in pavement performance modeling, by investigating the effect of different learning rates of adaptive logic networks in pavement performance modeling. Adaptive logic networks (ALN) has recently emerged as an effective alternative to artificial neural networks for machine learning tasks. Adaptive logic networks (ALN) is a network arranged as a binary tree, where the processing element has exactly two inputs and one output. Each processing element, or node, computes one of the four logical functions AND, OR, RIGHT (MAX) and LEFT (MIN). The root node of the tree represents the output of the network, while the leaves of the tree are connected to the input variables and /or their complements. It has been shown that such a tree can be constructed to compute any Boolean function for arbitrary number of input variables.