Study On Active Prevention And Fast Evacuation Control Of Urban Traffic Congestion

For traffic network,traffic congestion means the growth of travel time,the increase of living cost,lower quality of life.For urban traffic management,traffic congestion means that the road operating costs increase,the transportation network transport efficiency decrease,traffic accidents rates increase.From the point of the city development,traffic congestion means serious city energy waste,worsen air pollution and noise pollution.Research and solve the problem of urban traffic congestion has become an important issue of urban development and construction.From the point of view of traffic signal control to solve the problem of traffic congestion,not only to consider the reliability of the signal control strategy,but also to consider the initiative of the real time strategy.Considering the existing requirements of the signal control of city traffic congestion,conbining with the traffic state analysis of city roads and network traffic flow macroscopic simulation technics,this dissertation has researched and discussed in the following areas,including traffic state observation methods of road,intersection and network,network traffic flow macroscopic simulation model,different active prevention and fast evacuation strategies of traffic congestion in different control scops and traffic conditions,timing optimization model,the trigger timing and switch timing of different control strageties.The main scientific research and research results in this siddertation include:1.Traffic flow operation pattern of city roads is analysied on the basis of the basic theory of pattern ognization.Based on the characteristic of city traffic operation parameters and the fast forecasting principle of K-NN nonparameteric regression,the traffic short-term prediction model based on pattern recognition of traffic flow timing-series and the road traffic congestion observation method are established.Comprehensive observation of intersection traffic congestion is achieved through the establishment of intersection traffic state observation model.Using the standard deviation of the number of vehicles of all roads(SDVL)to characterize the spatial distribution characteristics of the vehicle density,the macroscopic fundamental diagram related to SDVL(SDVL-MFD)is proved to exist through SDVL changes analysis during the traffic deadlock formation process.Then the the macroscopic traffic state analysis method based on SDVL-MFD and the regional traffic state prediction method based on the SDVL change pattern recognition are proposed,which can make the active prevention and control of traffic congestion and lay a solid foundation.2.All possible traffic flow conflicts of an intersection is simulated by the traffic flow operation conditions on intersection-inside traffic flow paths,a novel macroscopic simulation model based on lane-group and intersection-inside traffic flow paths is proposed by modifying old macroscopic simulation model based on lane-group and is verified by model simulation analysis,and finally th city network traffic flow macroscopic simulation system is developed,which could provide fast simulation help to the traffic control problem.3.The general control stragtegies of intersectin signal control in different traffic conditions are analised and the calculation method of traffic delay time and queue length is studied in detail.The “discharge” signal strategy for intersection congestion active prevention in mild congestions and the “discharge + limiting” signal strategy for intersection congetion evacuation in moderate congestion conditions are developed,and related timing optimization model for these two control strategies are established.4.According to the characteritics of the traffic state when traffic bottleneck become active,regional traffic congestion active prevention control target and the trigger time are formulated.Based on the analysis of the correlation between adjacent traffic flows,the calculation model of the correlation of adjacent traffic flows is established.The regional congestion active prevention area division method based on the correlation between adjacent traffic flows is proposed,then a multiple-objective optimization model of congestion active prevention is set up,and the solving method based on macro traffic simulation model and improved Niche Genetic algorithm is put forward.5.Through the analysis of “external control and internal evacuation” theory,based on the calculation model of the correlation of adjacent traffic flows,the " internal evacuation " area(block area)and "external control" area(buffer area)classification method and the strategy trigger time are put forward.Through the analysis of the mutual influence of the traffic demand and signal control strategies of “internal evacuation” area and “external control” area,the optimization model of city regional congestion fast evacuation is established based on bilevel programing principle,and the model is solved using nested genetic algorithm.6.At last,the switching logic and trigger process of all the city traffic congestion active prevention control and fast evacuation strategies are cleared.The macroscopic simulation model proposed by this dissertation is used to simulate the network traffic operation of the Pearl River New Town of Guanghzou City.Using the proposed traffic state observation method based on traffic flow time-series pattern recognization,related traffic congestion active prevention and fast evacuation control method is triggered according to the traffic state observation results in real time to achieve the real time traffic signal optimization in congested conditions of Pearl River New Town.In general,a complete set of signal control theory and method of urban traffic congestion conditions is provided by this paper.