Influencing Long-Term Route Choice by Traffic Control Measures—A Model Study

Abstract

Currently used traffic control measures, such as traffic signals, variable speed limits, ramp metering installations etc., are often not designed to influence the route choice of drivers. However, traffic control measures do influence the travel times that are experienced in the network. Since route choice is, at least for a part, based on experienced travel times, the control measures thus also influence the long-term route choice. This influence can be seen as a side-effect of the control measures, but in this paper we will investigate the possibilities to explicitly and actively use the influence of the traffic control measures to change the long-term route choice. Using basic traffic flow and route choice models we investigate how outflow and speed limit control can affect the final equilibrium turning fractions. As an example we consider a case study for a simple network with two routes and use a simple linear outflow controller, which makes the analytical investigation of the effects of the controller possible, but the results can be extended to more sophisticated control methods.

Control of day-to-day route choice in traffic networks with overlapping routes

“…We have developed a control approach based on model predictive control to influence the route choice of drivers, using existing control measures like outflow limits or variable speed limits. In previous work (van den Berg et al (2008b(van den Berg et al ( ,a, 2009) we have considered route choice control for basic networks and constant or piecewise constant demands. In this paper we have extended the model to include networks with overlapping routes and with restricted link inflow capacities.…”

Day-to-day route choice control in traffic networks with time-varying demand profiles

“…This research is a continuation of the work described in [10], [11] where we have considered a constant demand. In this paper, we extend these results to time-varying demand profiles.…”

“…In this section we extend the day-to-day route choice model of [10], [11] to the case of time-varying demand profiles, and we present the corresponding MPC-based route choice control approach.…”

“…More specifically, the free-flow travel time at day d along route r is given by: τ free r (d) = l r v r (d) . The model of [10] for the average experienced travel time assumes that the travel time τ r on a route has two components: the time spent in the queue τ queue r and the freeflow travel time τ free r :…”

“…Route choice model: Route choice models describe the route choice of drivers at locations where a route must be selected. The model of [10] updates the turning rates for the next day d + 1 based on the difference in travel times on the current day d between the two routes, while also taking into account that the turning rates are bounded between 0 and 1:…”