Day-to-Day Traffic Evolution Model And Week-to-Week Intervention Strategies For Bi-Modal Networks

The urban transport network is often in a non-equilibrium state due to the constant interferences and the changes of internal and external factors.Therefore,the traditional equilibrium analysis theory would not apply to the problem analysis.The day-to-day dynamic traffic modeling method is widely adopted to accommodate the limitations of conventional methods.However,most of the existing studies in the day-to-day dynamic traffic modeling framework focus on single-mode networks.Only a small number of researchers adopted dual/multi-model traffic networks,where strict assumptions are often imposed.Moreover,the related management measures also still need to be strengthened.For these reasons,this thesis is interested in the bi-mode(i.e.,cars and buses)mixed transportation system.We carefully investigate the differences in operating characteristics and the passenger experience between the two modes and then establish a more reasonable day-to-day traffic evolution model.After that,we further explore more reasonable traffic pricing and subsidy strategies to reduce congestion and improve the environment.These research efforts can not only deepen our understanding of the operational laws and various complex phenomena of the bi-modal networks.A more reasonable theoretical support for realistic decision-making is also expected.The main contents of this thesis are summarized as follows:First,by systematically considering the travelers’ distinct perception of the two traffic modes,the in-vehicle congestion,and the individual travel choice behaviors,we establish a day-to-day dynamic adjustment process model to capture the aggregate individual travel decision-makings(including travel mode and path)underlying the macroscopic network traffic evolution on a car-bus bi-modal transport network where dedicated and mixed sections coexist.After that,numerical analyses are conducted for the constructed model.Second,based on the established day-to-day dynamic model,a week-to-week dynamic pricing model with the objective of alleviating congestion and reducing emission is established for the bi-modal network.This model systematically considers the pricing acceptability,stability,and individual travel adjustment behaviors,etc.Considering the model’s poor mathematical properties,a pattern search algorithm incorporating the simulated annealing component is designed to solve it.The algorithm and model are further analyzed through numerical examples.The experimental results initially verify the rationality of the model and the effectiveness of the algorithm.Last,based on the model in the second step,we further consider using the pricing to subsidize the bus fare and establish a dynamic and revenue-sustainable pricing-subsidy model with the goal of improving the efficiency and environment.The model still solves the revised pattern search algorithm developed in the second step.Numerical studies are performed to analyze the model and compare the efficiency of various management strategies.