Study On The Vulnerability Of Road Network Traffic System Considering The Characteristics Of Congestion Propagation

In recent years,domestic and foreign researchers have carried out extensive research on road congestion and vulnerability for road network based on different theoretical methods and models,but so far the definition of vulnerability of road network has not formed a definite and unified concept.Especially,due to the complex factors,such as road structure,extreme weather,traffic accidents,etc.,the traffic congestion source may shows the dynamic characteristics which will affect the space-time distribution of traffic flow,and even lead to the traffic guidance capacity of trunk roads,branch roads and the unobstructed capacity of local road network become hard to control,then the running state of road network in different roads and local areas shows different degree of vulnerability.In order to study the impact of congestion propagation on the vulnerability of urban road network,an identification model of vulnerability for urban road network is established by analyzing the evolution law of traffic congestion.The main work of this dissertation is supported by National Science Foundation of China P.R.(NSFC)under grants 61573706 and 61703063,the Scientific Research Foundation for the Returned Overseas Chinese Scholars under Grant 2015-49,etc.The main work and innovations of this dissertation as follows.(1)A new forecasting method for short-term traffic flow with event-triggered strong fluctuation is proposed.Firstly,directing against the characteristics of short-term traffic flow with strong fluctuation caused by the vulnerability of road network under the impact of traffic events,a novel improved grey prediction model is constructed based on the traditional grey model and Markova chain.Secondly,combined with the trend prediction advantages of the first-order difference exponential smoothing algorithm,a new effective short-term traffic flow prediction method was constructed by introducing the dynamic weighting factor.The experimental results show that this prediction method can effectively predict the short-term traffic flow with event-triggered strong fluctuation,and can effectively manage and guarantee the overall capacity of the road network macroscopically.(2)A method of identification and prediction of urban traffic congestion via cyber-physical link optimization is proposed.Firstly,aiming at the parameter optimization of LiESN,the differential evolution algorithm(DEA)is used to calibrate the key parameters.Secondly,the road weight allocation is carried out by Pearson correlation analysis considering the specific topology characteristics of local road network.On this basis,the corresponding flow-frameworks are designed.Then,the rationality and validity of this method are verified by using the congestion delay data set of local road network in Chongqing Nan'an district,which is acquired from AutoNavi Maps.(3)An assessment method of vulnerability for road network based on cloud model is proposed.Firstly,on the basis of topological mapping of road network structure of Nan'an distinct by dual method,the structure vulnerability of road network structure is studied by taking degree and network efficiency as the main evaluation indexes.Secondly,in order to describe the state of road congestion,cloud model theory is introduced to characterize the characteristics of road congestion.On this basis,a state vulnerability evaluation model based on congestion cloud graph is constructed.Finally,experiment and analysis are carried out in the light of actual congestion delay data set of Nan'an district provided by AutoNavi Maps.The results show that the model can effectively describe the vulnerability of road network.