Continuous Network Design Problem Considering Stochastic Demand

Most existing network design models ignore uncertainties that usually have negative impacts on the performance of a road network. A bi-level model for the continuous network design problem (CNDP) is proposed, in which uncertainties due to the normal day-to-day variations in the travel demand are explicitly considered. At the upper level problem, the system performance index is defined as the weighted sum of the expected total travel time and the investment costs of link capacity expansions. At the lower level, user equilibrium is used to simulate the travelers' route choice behaviors under stochastic demand. Then a simulation-based genetic algorithm is presented for the model. Finally a simple numerical example is provided to illustrate the model and the algorithm.