A Tri-objective Resource Allocation Model for Pavement Rehabilitation by Hybridization of Parametric and Constraint Methods

One of the important issues in road pavement management is the allocation of resource in different parts of the network in order to maintain the pavement conditions at desired levels. In this paper, a tri-objective model is proposed for this purpose, whose objective functions are: 1) to minimize the percentages of the network that are in “critical” and worse-critical conditions (other than the “worst” condition), on which some improvement measures should be performed to reach a desired condition; (2) to minimize the percentage of the network that is in “worst” condition, and, in order to bring it to a desired condition, it should take renewation measures; and (3) to minimize the total cost incurred due to the maintenance and rehabilitation measures made during the entire planning period. To solve the tri-objective model, a hybrid method is proposed using multi-objective parametric optimization and ε-constraint methods. The results of applying the model show that despite the mental imagination at first glance, the first and second objective functions may have conflit, so that increasing one of them may lead to decreasing the other; therefore, these two objectives cannot be considered together as a single objective function. One of the important features of the proposed method is that the Markov Chain process model, as the prediction tool for the pavement condition, is incorporated into the optimization model. This makes the pavement condition prediction and resource allocation simultaniously and prevents achieving local optimal solutions.