Urban Road Network Design Problem Model For Multiuser Transportation System With Considering The Intersection Delay

Base on the analysis of the current researches of the road Network Design Problems (NDP), this thesis aims to optimize the algorithm which used to solve the urban road network design problem. For the previous models, the lower model in the NDP model mostly considered single user in the road network. But the real is that different user intent to choose different way to finish a journey. This thesis builds an assignment model which considered multi-users divided by income. Meanwhile, the form of the road impedance function in the traffic assignment model is improved and designed an algorithm to get parameter values. A bi-level programming model is established to solve the Continuous Network Design Problem (CNDP). The internal circulation of the model is to solve the travel demand-based assignment for multimodal and multiuser transportation system, while the external circulation is to solve the road network design problem.The main works of this thesis are as follows:Established travel demand-based assignment model for multimodal and multiuser transportation system. The impedance function was improved with considering the effect among different modes in the same road. Users tend to choose different modes of transport into a multi-user category according to their income. And then a two-tier network structure was proposed to describe such a system, in which the first-tier network is used to depict the traveler's mode choice behavior and the second-tier network is used to depict the vehicle routing when a certain mode has been selected.Construction and solving traffic assignment models with considering the intersection delay. Based on traditional user equilibrium traffic assignment model, the intersection delay was added into the model. And the data storage structure was optimized, and then designed the Dijkstra algorithm which was used in the Frank-Wolfe algorithm.A bi-level programming model was established to solve the CNDP problem and the algorithm to solve the problem was improved. The lower level is the multi-modal and multi-user transportation system which considers the intersection delay, while the upper level is constrained by the total travel time of the road network and the government investment. Improved the traditional Particle Swarm Optimization (PSO) algorithm and joined the natural selection in the iterative process to improve the computational efficiency.An example was used to validate the model and explore the sensitivity of the main parameters. The Sioux Falls was selected as the example in order to further test and verify the feasibility of the algorithm. Then the travel demand of different users was quantitative analyzed. Finally, combining the influence of the main parameters in the iteration, the thesis analyzed the sensitivity of the parameters which could provide a reference for the practical application of the model.