The increasing age of national highway bridge stocks demands novel management policies that can maintain or increase safety levels in the light of reducing and budgets from the government. The French Highways Agency uses its Image Qualité Ouvrages d'Art – Structural Quality Imaging (IQOA) inspection programme to determine the conditions of its bridges. The present paper principally describes the development of a model to predict different trends in bridge condition rating in the light of different management strategies. Whilst the approach described is intended to be a general method for managing bridges, it will be demonstrated by particular reference to the stock of reinforced concrete bridges managed by the French Highways Agency. The prediction model utilises Markov chains, which were chosen because they allow a non-parametric discrete model to be built, to evaluate the varying condition of the whole bridge stock over time. The development of the model is described in detail, together with how the hypotheses for applying the Markov chains were validated to accurately predict the transition matrices between different IQOA states. The paper then goes on to describe how the model is used to differentiate between different management strategies, and to predict how future costs can be optimised to give the most efficient management actions across the whole bridge stock under different constraints.
[1]
R. Frank Carmichael,et al.
Microcomputer Bridge Management System
,
1993
.
[2]
John Odentrantz,et al.
Markov Chains: Gibbs Fields, Monte Carlo Simulation, and Queues
,
2000,
Technometrics.
[3]
Mingxiang Jiang,et al.
Modeling of risk-based inspection, maintenance and life-cycle cost with partially observable Markov decision processes
,
2005
.
[4]
Erik H. Vanmarcke,et al.
Modeling Bridge Deterioration with Markov Chains
,
1992
.
[5]
Hid N. Grouni,et al.
Methods of structural safety
,
1986
.
[6]
Changqin Liu,et al.
Maintenance Strategy Optimization of Bridge Decks Using Genetic Algorithm
,
1997
.
[7]
William T. Scherer,et al.
Markovian Models for Bridge Maintenance Management
,
1994
.
[8]
Ross B. Corotis,et al.
Optimal Life-Cycle Costing with Partial Observability
,
2000
.
[9]
G. Morcousa,et al.
Maintenance optimization of infrastructure networks using genetic algorithms
,
2004
.
[10]
C Cremona,et al.
Optimization of reinforced concrete bridges maintenance by Markov chains
,
2006
.