01026051

Component

This paper presents a methodology wherein Markov transition probabilities, Genetic Algorithm (GA) and Monte Carlo simulation have been used to address the issue of optimum decision making and uncertainty analysis at the programming level of Pavement Management Systems (PMS). Markov transition probabilities have been used to forecast the pavement network condition at the programming level and uncertainty analysis in this paper refers to the impact of uncertainty in the transition probabilities on the projected outcome of the decision. It is known that simultaneous run of GA might produce results that are slightly different. This property of GA has been explored in this paper to derive not just one but more than one optimal solutions. Finally Monte Carlo simulation is used for the uncertainty analysis of each optimal solution obtained from GA. A hypothetical example has been presented at the end to demonstrate the proposed methodology. Keywords: pavement management systems, programming level, Markov transition probabilities, optimization, genetic algorithm, uncertainty analysis, Monte Carlo simulation

01020180

06-2944

English

21p

2006

Transportation Research Board 85th Annual Meeting

CD-ROM

Figures (8) ; References (19) ; Tables (2)

Decision making; Genetic algorithms; Monte Carlo method; Optimization

Pavement Management Information System

Markov transition probability

Design; Highways; Pavements; I22: Design of Pavements, Railways and Guideways

Transportation Research Board Annual Meeting 2006 Paper #06-2944

TRIS, TRB

Mar 3 2006 11:10AM