An Agent-Based Framework for Infrastructure Maintenance Decision Making
Abstract
A transportation system plays a significant role in the economic development of a region by facilitating the movement of goods and services. No matter how well the infrastructure is designed or constructed, it is beneficial to know maintenance needs in the life cycle of the infrastructure so that the service life of the pavement is prolonged by minimizing its life-cycle cost requirements. The pavement life-cycle performance helps maintenance investment decisionmakers to efficiently utilize the available infrastructure funds. This research focuses on developing a framework for identifying a more effective and efficient way of decision making on the management and maintenance of infrastructure. The developed framework uses an agent-based modeling approach to capture the interaction that exists between different components of the transportation system and their characteristics such as traffic volume and pavement condition, user cost, agency cost, etc. The developed agent-based is useful to investigate the effects of timevarying vehicular density on pavement deterioration and the road users’ driving behavior. The developed framework was demonstrated as a two-lane highway as a case study. Using the developed agent-based simulation framework, it was possible to identify when the road infrastructure maintenance should be done to increase the desired PCR value. Also, it was possible to show a decrease in PCR can affect the cost of road users. The framework can track the time to determine when maintenance should be done based on the PCR values that determine whether the pavement is in good, moderate, or bad condition. Regardless of the degree of road users’ patience to stay in their travel lanes (patience), the vehicle distribution on the road is balanced in the long run because road users tend to change their travel lanes to minimize their overall travel times. When the patient level is low, road users tend to change lanes more, causing a high number of vehicles on the left lane as it is considered the lane changing lane in two-lane highways. It was also observed that as the patience of the road users increases, the number of vehicles that use the right lane is almost the same as the number of vehicles that use the left lane.
Degree
M.Sc.
Advisors
Woldemariam, Purdue University.
Subject Area
Computer science|Economics|Energy|Management|Recreation|Systems science|Transportation
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.