Reliability-Based Marginal Cost Pricing with Stochastic Demand

The paper explores the marginal cost pricing (MCP) issues in a traffic network with stochastic demand. It has been known that the naïve method by simply replacing the link flow and link travel time in the original MCP with their expected values is not workable. While on the other hand, when network uncertainty is accounted for, not only the mean travel time but also the variance of travel time, i.e. the reliability, are cared by travelers. In this study, the value of reliability is incorporated in both the travelers´ route choice criteria and the objective of system optimum. Then, a reliability-based MCP, which drives a user equilibrium flow distribution under stochastic network toward system optimum, is presented. A numerical example is examined in different levels of uncertainty and with different values of reliability. It exhibits that the higher the uncertainty and the value of reliability, the worse the naïve method.