1367602 - Network Pavement Life-Cycle Cost Optimization: A Piecewise-Linearized Approach

With 76% of goods moved by trucks, millions of trucks traverse the highway infrastructure annually. This requires optimal planning of pavement preservation, maintenance, and rehabilitation to ensure efficient and timely delivery of goods. Assigning a budget that is based on real-life pavement condition that includes both agency and user costs (e.g., fuel, work-zone, maintenance, and tire wear tear) is crucial for agency budgeting. Because of its nonlinearity, it has always been addressed either by using metaheuristics (e.g., genetic algorithms), which don’t give a guarantee on optimality or by using assumptions that simplify it to assumptions that simplify the problem a lot to an extent in which it can be dealt with as a linear program. In this study, the (piecewise) linearization of various components. The international roughness index (IRI) progression, an important parameter in calculating user cost, was piecewise-linearized to keep the shape of the function (an R2 of more than 0.99 could be achieved). Other components were linearized. This was achieved by a novel formulation that kept the problem (integer) linear. The planning horizon was assumed 20 years and a budget was assigned for that period. Bid lettings from the state of Illinois were used as cost data. The model is used to create charts of average optimal IRI for different network conditions (e.g., network size, traffic, etc.).