Urban Transportation Demand Management Of Multi-period, Multi-user Sub-optimal Congestion Pricing Model

Rapid developments in urbanization and growth of car ownership in China have led to serious urban traffic congestion. Conventional supply-side strategies such as road capacity expansion have been less effective in reducing traffic congestion. More and more transportation policy-makers are now realizing the importance of transportation demand management (TDM). Congestion pricing as a TDM strategy has received increasing attention. This paper proposes a second best congestion pricing model with multi-period in the form of bi-level programming. Besides, by incorporating user dimension together with the spatial and temporal dimension, the above model becomes a second best congestion pricing model with multi-period and multi-user. The models with equity constraints are also studied. After analyzing the optimal conditions and congestion tolls of the models, a simulated annealing method is applied for the bi-level programming problem and demonstrated with a simple urban network example of two-period and two-user. The simulation results show that both the model and the algorithm are effective.Generally speaking, major impacts of congestion pricing policy on traffic volume allocations include: diversion of traffic from the tolled links to the free links;shift of the peak period traffic to the off-peak period;reduction in total traffic volumes. Congestion tolls which are endogenously determined by the equilibrium traffic volume allocations can actually be regarded as results of the game between the government and users. Congestion tolls as well as the social welfare are closely related to the second-best network toll design, such as the tolled link, period and users etc. It is true that social welfare gains through congestion pricing, however, the traffic congestion may not be alleviated. Tolling in peak period or encourage more users of high income to travel in peak period are more effective to prompt the social welfare. It is inevitably to bring inequity for congestion pricing and equity conflicts with social welfare each other. The results show that when equity constraints are released to a certain degree, continuing releasing the equity constraints won't lead to social welfare gain, while releasing spatial and temporal constraints may be more effective.