A route guidance system with personalized rerouting for reducing traveling time of vehicles in urban areas

Urban road traffic congestion has been a severe problem for decades. There are many route guidance systems (RGSs) providing rerouting suggestions for drivers to avoid on-going congestion. However, these systems use macroscopic approaches to generate flow-level route guidance and provide homogeneous guidance to all drivers that belong to the same flow, which may simply shift the on-going congestion from one point to another. In order to overcome the demerit of existing RGSs, we propose a RGS with a microscopic rerouting strategy to generate vehicle-level personalized route guidance. The mechanism of the proposed RGS is to rank the vehicles that need to be rerouted according to a certain criterion, and compute detour route sequentially for each vehicle based on effective link travel time (ETT). The ETT is updated every time a vehicle is rerouted to account for the potential impact of this vehicle on the following rerouting. Simulation studies show that with proper guidance update interval and ranking criterion, the proposed RGS successfully reduces the average travel time by 22% compared to the existing one. Besides, 40% less vehicles are involved in rerouting in the proposed RGS, and the average number of reroutings for each rerouted vehicles is reduced from 6.8 to only 1.6. Performance analysis on microscopic level suggests that the advantages of the proposed RGS may primarily come from the vehicles that are rerouted in both RGS and the vehicles that are rerouted in the existing RGSs but not in the proposed one, which in total takes up 77% of all the vehicles. The proposed RGS also left-shifts the distribution of travel time so that the complied drivers have high possibility of reducing travel time. We conclude that the proposed RGS with personalized rerouting strategy can not only improve the operation efficiency of a traffic network, but also give incentive to drivers to follow the guidance that they receive.