Study On The Framework And Key Technologies Of Macro Simulation System For Urban Transportation

Transportation planning restricts a city's future, and macroscopic traffic simulation system is the fundamental means to design and evaluate transportation planning programs. In this dissertation, macro simulation system of urban transportation is the theme, and the framework and key technologies of macro simulation system which is based on 4-step model, are studied in line with China's national conditions and operability. In order to overcome the shortages of current technologies, some creative and useful methods and algorithms are proposed. Main studies and results of this dissertation are as the following:1. Macro simulation system framework of urban transportation in conformity with national conditions is constructed. In this dissertation, simulation system frame is designed based on requirement analyses, and the functions of the simulation system are subdivided based on 4-step model. Then, an urban traffic simulation architecture which contains 12 sub-modules is set up, and this architecture is universal and adapts to the requirements of China's current conditions very well.2. Improved OD metric calibration model——3D balancing gradient searching method is put forward, and OD metric calibration procedure and test standard are proposed in line with the actual situation of domestic cities. In this dissertation,3D balancing gradient searching method is advanced to overcome the shortages of general balancing gradient search method. Practice has shown that the proposed method can improve the OD metric calibration accuracy effectively. Furthermore, after analyzing the convergence, we raise the operable procedure of OD metric calibration, and GEH test standard which refer Guangzhou's survey practice and dada.3. In view of transportation requirement, sampling expansion for resident trip survey, and time value analysis model based on SP survey are proposed. For resident trip survey this dissertation raises the "family-vehicle-family" sampling expansion method, which reduces the deviation of the direct sampling expansion method effectively, reflects the implied transportation requirements very well, and makes the results more reliable and reasonable. Moreover, for improving the forecast accuracy of trip distribution and mode, we analyze the principle of SP survey and put forward the test and survey design program, discuss trip time factors, and builds up time value Logit model.4. Traffic equilibrium assignment model based on the knowledge of the road network, and the solving of the model are discussed. This dissertation analyzes the deviation of UE model and SUE model in real applications, proposes a new traffic equilibrium assignment model in accordance with people's knowledge of the road network, and the solving of this model. The results of the computational example show that this new model was more elastic than UE model and SUE model in non-crowded traffic state, and the new model could be considered as a realistic method for traffic assignment.5. We introduce the design and realization of Guangzhou's comprehensive traffic simulation system, and discuss the application of this system in real city transportation planning. Based on the study of the architecture and key technologies of macroscopic traffic simulation system, this dissertation presents the design and realization of Guangzhou's comprehensive traffic simulation system in detail. Furthermore, the system's applications, such as Guangzhou's urban passenger transport development forecast, Guangzhou's rail network planning and so on, are analyzed.