Linear programming models for a fair and efficient traffic assignment in congested road networks

Solving the road congestion problem is one of the major issues in modern cities since it causes time wasting, pollution, higher industrial costs and huge road maintenance costs. Among possible congestion avoidance methods, traffic assignment is a valuable choice since it does not involve huge investments to expand the road network. Traffic assignments are traditionally devoted to two main perspectives on which the well-known Wardropian principles are inspired : the user equilibrium (user's perspective) and the system optimum (system perspective). User equilibrium is a user-driven traffic assignment in which each user chooses the most convenient path selfishly. It guarantees that fairness among users is respected since, when the equilibrium is reached, all users sharing the same origin and destination will experience the same travel time. The main drawback in a user equilibrium is that the system total travel time is not minimized. On the other hand, the system optimum is a system-wide traffic assignment in which drivers are routed on the network in such a way the total travel time is minimized but users might experience travel times that are higher than the other users travelling from the same origin to the same destination. Thus, there are drawbacks in using one of the two assignments that can be partially overcome by applying users' fairness considerations while minimizing a system-oriented objective. In the last decade, few attempts have been done to present a users' needs and system efficiency trade-off traffic assignment with non-linear programming techniques. In this thesis linear programming models for a fair and efficient traffic assignment in congested road networks are presented.