Markov-Based Bridge Deterioration Model in Texas.

摘要

In Bridge Management Systems, bridge Network-level models are used to manage the population of bridges in a certain area. Due to limited funding and the increasing number of aging bridges, a good bridge network-level deterioration prediction methodology is needed. To provide information to managing bridges, the Texas Department of Transportation (TxDOT) has been inspecting bridge condition using discrete bridge elements since the year 2000. The TxDOT historical elemental inspection information was summarized by this thesis on an element by element basis for data spanning inspections from 2000 to 2008. A methodology to estimate the parameters of a Markov chain based bridge deterioration model was then developed and is documented by this thesis. Error minimization routines where also summarized in order to automate the estimation of the Markovian transition matrix parameters for each bridge element in the Texas bridge network. The recommended bridge deterioration modeling procedures in this thesis will support key bridge management decisions for the Texas bridge network, by allowing future development of routines to predict the condition and maintenance needs of the Texas bridge network.