Research On Combination Optimization Models Of Congested Road And Parking Pricing

Parking charge(PC) and congestion pricing(CP) play an important role in the implementation of Travel Demand Management measures. Due to the charges, individual trip choices and traffic demand distribution are influenced. In addition, PCCP can be available to decrease the traffic flow density of congested areas or roads. However, there are difficulties in the CP practice, and PC is second-best for relieving traffic congestion. The attention is focused on the strategy combining PC with charges on certain roads in this thesis. Especially the pricing problem is mainly analyzed.Allowing for suburbanites with the purpose of going downtown, they can drive cars from origin to destination or transfer to a bus after travelling a distance. Analyses of travel characteristics, including trip cost, mode choice, path selection and transfer stop choice, are carried out in the two-mode traffic network. Then an optimization model for combination charging is built under the condition of mixed stochastic user equilibrium traffic assignment. The given model is solved by chaotic simulated annealing algorithm in a simplified traffic network. The results show that the combination charging strategy not only greatly reduces private traffic flow, but also motivates more travelers to choose the combined travel mode.Considering differences of value of time(VOT) among travelers in actual, travelers are bracketed into different groups based on VOT. Together with influences of PCCP to traffic demand and trip cost, a bi-level programming model is given to describe the combination charging strategy for multi-class users under elastic demand. The upper-level object is maximizing the social benefit, and the lower-level programming problem is the multi-class stochastic user equilibrium model with elastic demand. The numerical example validates that combination charging has the superiority to alleviate congestion and curb traffic demand. Furthermore, the attitudes of different class users towards combination charging can be reflected by the model.