Optimization of Network Pavement Life-Cycle Cost: A Piecewise Linearized Approach

Abstract

Optimal management of pavement assets becomes important because of the escalating challenges in this field. Managing the network of paved roadways in the United States necessitates the introduction of optimization tools, such as mathematical optimization. Although several efficient optimization techniques are available, an effective approach requires a special structure of the problem. The optimization of a pavement maintenance and rehabilitation schedule poses a complex challenge, primarily because of two factors: nonlinearity and the presence of integer decision variables. Nonlinearity exists as a result of multiple sources. One source is pavement condition, commonly measured by pavement roughness. This study introduces a method that uses piecewise linearization of the pavement roughness progression function. Circular shift was used to linearize the resulting optimization model. A hypothetical city, the size of Cook County in Chicago, was used as a case study. Both agency cost and user cost were considered in the model. Agency cost was determined from consultations with professionals and online data, whereas user data on existing models. The study demonstrated that increasing the agency cost by investing one dollar per lane mile per year has a high return on investment until a certain threshold, beyond which allocating more budget does not lead to a reduction in life-cycle cost.