Ant Colony Algorithm for Modeling Planning of Maintenance Management in Road Networks

Every country's road transportation network plays a vital role in their economy, which is evidence for the great importance of the physical condition of infrastructure. The primary goal of this research is to introduce an optimized program for the timing of the maintenance and rehabilitation of the network's pavement sections. To this end, a non-linear programming model was applied alongside the Max-Min Ant System Algorithm in order to select an optimized set of road network sections, over a five-year planning horizon. The Markovian Prediction Model was also employed to predict each pavement section's transformation over time or after any maintenance and rehabilitation operation. The Max-Min Ant System Algorithm was used in order to enhance user satisfaction and minimize the M&R costs in the course of the planning horizon. The model's constraints included the allocated budget for the planning period and the treatment length per year. The proposed model was implemented on a hypothetical network of 40 sections, and the optimized timing for the maintenance and rehabilitation of the road network was presented. Furthermore, the budget sensitivity analysis results indicated that a budget increase of up to 50 percent of the M&R costs of the whole network can improve the objective function, but a further increase will not yield a significant benefit on the objective function or public interest.