Service level-oriented route guidance for overlapping routes in road networks: A comparison with MPC

Service level control is a promising strategy to operationalize policy objectives within road networks. In previous work we have introduced a route guidance approach that controls service levels over two alternative routes. However, when applying the method on a network level with multiple actuators, controlled routes might overlap. Interaction between the controllers then occurs, because the traffic that flows towards overlapping route stretches can be manipulated from multiple directions. In this contribution the interaction mechanism is described and verified by means of a simulation test case. The functioning of the proposed method is compared with a model predictive route guidance controller that realizes optimal conditions given the prevailing policy objectives. Results show that the interaction mechanism successfully prevents network performance degradation by directing traffic such that available capacity over route alternatives is fully utilized. If target service levels are properly defined, queue spill back towards upstream infrastructure is delayed or even prevented. With respect to equity, the performance difference over the routes is limited by degrading and recovering the target service levels of the controlled routes stepwise. To conclude, the approach can be tuned such that optimal performance is approximated, but with a significantly lower computational demand and a better scalability with growing network sizes.