Detecting the warning level of forest roads pavement using the genetic algorithm

Assessment of forest road pavement condition and its optimal cost management is backbone of a forest road network. In addition, an optimum budget allocation is a challenge during road life cycle. The precise assessment requires special equipment that is often neglected due to its absence or the associated high expenses. Therefore, it is recommended to use techniques that can provide cheap and fast linear and nonlinear modeling of pavement deterioration. The aim of this study was to investigate the performance of Genetic Algorithm (GA) to detect the warning level of forest road pavement as well as to plan the pavement maintenance. The GA was evaluated to identify the warning level of pavement and provide a plan for the maintenance of the road network over a period of five years based on annual plans across a test site of 5 series located 50 km west of Haraz forest road network. The results showed the optimum protrusions warning level for five centimeters damage and rutting 10 cm. Moreover, it was shown that an optimum planning combined with regular repair and maintenance work enables reducing the associated costs from 25 to 73 percent during the first years. It showed the importance of regular maintenance activities at the network level of pavement and application of novel techniques to allocate budgets for pavement maintenance.