Article ID: | iaor20171812 |
Volume: | 17 |
Issue: | 1 |
Start Page Number: | 313 |
End Page Number: | 337 |
Publication Date: | Apr 2017 |
Journal: | Operational Research |
Authors: | Yu Bin, Yao Baozhen, Zhang Liu, Guan Feng, Peng Zixuan |
Keywords: | networks: flow, networks, combinatorial optimization, simulation, transportation: general, demand, search, heuristics |
Congestion pricing is considered as an effective measure to ease urban traffic congestion. This paper develops a bi‐level model that designs more equitable congestion pricing with bus subsidies. In the upper‐level model, the constraint of alternative path is proposed to ensure that there is at least one non‐toll path for any origin–destination (OD) pair. If one link is the only way for an OD pair just like Lidingö, Stockholm, this link should not be tolled. Meanwhile, toll revenue is used to improve the service level of transit system in which transfer fare is subsidized and the frequency of the most crowded bus is improved. The lower‐level model is a multimodal transportation network equilibrium model. A shuffled pattern search algorithm and method of successive averages are used to optimize congestion pricing, transfer fare and bus frequency in the bi‐level model. The results of the numerical test show that the model can meet the travel demand of travelers after tolling. Moreover, traffic congestion is alleviated while service level of transit is improved.