The Research Of Grid-based Urban Traffic Flow Simulation

With the development of economy and the acceleration of urbanization, the numbers of automobiles in cities and demanding of transportation have been sharply multiplying. The traffic jam becomes more and more serious which causes great waste in traveling time, great increase in traffic accident and great deterioration in environment. To a great extent, it restricts the high-speed development of country's economy. To advance a scientific and effective solution for city traffic jam which also can be applied in city traffic programming and managing, this thesis does research on the key technologies in parallel simulation considering the characteristics of the macroscopical traffic simulation and designs Grid-based Urban Traffic Flow Simulation (GUTS) System in MPI environment on Globus flatform.The thesis includes following contents:(1) It analyses on the possibility, reliability and necessity of building GUTS system.(2) It theoretically analyses on the GUTS system and designs in structurization of functional modules in the system. It analyses five functional modules of the system and structures a model of simulating traffic system.(3) It designs the GUTS by an object-oriented method.(4) It analyses and predicts OD data, and advances a grid-based optimum solution or method for service of route choice by combining with the variable OD data distributions of traffic information in road web.(5) It analyses the parallel calculating task scheduling in different environment, and two scheduling algorithms (PMM and BM) for the heterogeneous computing environment are given.(6) It analyses the synchronize strategy of parallel simulation. Aiming at the synchronize strategy, it also analyses advantages and disadvantages of existing four kinds of synchronizing strategies, and gives a new synchronizing solution.The GUTS can help travelers reach the destination along the best path from the start to the end, and it can effectively avoid traffic jam and decrease the time delay of vehicles on the road. What's more, it can make road web more clearly and more efficient to realize the balance of traffic flux in whole web.