| With population growth, rapid development of the national economy and promotion of urbanization, more and more cars appeared in urban cities. As a result, the contraction between traffic supply and demand becomes more and more prominent in large and medium cities, leading to severe traffic congestion. This has a bad effect on the cities and restricts their developments. Consequently, how to solve the traffic congestion has become the key point of urban sustainable development.The method of pre-diversion for urban center district is proposed in this thesis. Prediction, pre-diversion and evaluation of traffic congestion are all included in this scheme. Traffic information is mainly obtained by floating car system and loop Real-time Acquisition system. However, this can also be realized by other technology:Internet of things, Video Fuzzy Recognition model and so on. Firstly, predict the traffic congestion. Discrete analysis is done basing on the historical data collected by floating car system, which can be used for macroscopic prediction of traffic congestion of road net and traffic congestion of rush hour. As to the microscopic prediction, small step method should be also used to predict the traffic congestion degree of certain road section. Secondly, evaluate the situation of traffic congestion by using the congestion index of road section and road net. Besides, the congestion index of road section is used to estimate the traffic congestion degree of road section while the other one is used to estimate the comprehensive congestion degree of road net. Finally, pre-diversion of traffic congestion is the key point of the thesis. The concrete pre-diversion plan is as follows:define the goal flow and diversion point and then work out the optimal route for passengers in order to reduce the travel time. |