Evaluation Model For Urban Road Network Vulnerability Based On Macroscopic Fundamental Diagrams

Due to the contradiction of supply and demand of urban traffic system,meticulous governance for urban traffic is required in further level by the high efficiency of urban lifestyle.The service level of urban road network would be badly affected by adverse events,which brings challenges to urban traffic network vulnerability evaluation and meticulous governance.Urban road network vulnerability analysis with the consideration of time-varying network operation states and traffic flow has become an important issue in the field of urban traffic management.Thus,strengthening traffic flow analysis under vulnerability conditions and systematic study of evaluation and responses for urban road network vulnerability are needed,which is important to advance urban road network vulnerability and mitigate congestion.The theory of macroscopic fundamental diagram(MFD)is used to develop insights into road network vulnerability with the considerations of dynamic vehicular flows.Systematic vulnerability analyses based on macroscopic fundamental diagrams are proposed,including dyanmic models of MFD for time-varying macroscopic road network operation states,vulnerability evaluation and procative responses strategies.In terms of urban road network vulnerability with structural negative consequences(i.e.,topology disruptions),this study proposes a novel topic and discusses how to analyze urban road network vulnerability without structural negative consequences under the condition of time-varying network operation states.According to vulnerability analysis with static traffic assignment,this study develops a methodological framework to investigate vulnerability based on MFDs and dynamic vehicular flows.Vulnerability results are measured by MFDs related to road capacity,which compares differences of two upper bounds of MFDs of normal network and abnormal network.Considering dynamics of transportation system and its patterns of MFDs,this study investigates time-varying characteristics of network operation states and converts dynamic behaviors of network states represented by ordered pairs in an MFD pattern into vectors on a coordinate system.The dynamic behaviors of all network states from the first point to the last point in an MFD pattern can be formulated by a substitution matrix.Then hysteresis from the perspective of traffic dynamics are characterized.A relative degree of decoupling of network average density variation from network average flow variation is proposed to measure the decoupling relationships over a specified time interval.According to accurate calculations of circulating vehicles,this study proposes an MFD estimation method based on automatic license plate recognition data(ALPRD)and a data fusion method that separates the network into two sub-networks,one with loop detectors and one with cameras.Empirical observation results from both ALPRD and fusion results show that arterial networks have coupling,decoupling and recoupling relationships between network average density and flow in rush hour.To measure the differences of MFDs of degradable urban road network,a well-defined regular expression of an MFD based on set theory is proposed,which defines the ordered pairs on a graphical MFD as elementary events.Then we focus on investigating the proposed governing bound of MFDs which is an invariant upper bound of MFDs of an established transportation system.The equivalence of invariant upper bound of MFDs and network equilibrium states is verified.A straight-forward programming-based method is constructed to derive invariant upper bound of MFDs under equilibrium conditions.The invariant upper bound of West Jordan network is derived by equilibrium network data.Based on geometric transformation of MFD,this study proposes a curve approximation problem of invariant upper bound of MFDs.An interpolation approach is used to solve the curve approximation problem.Then a quick query list for transportation network equilibrium states is calculated.For road network vulnerability caused by negative incidents,a general network vulnerability index based on MFD is proposed.Then a case study in West Jordan network which was usually threatened by regular heavy snow during wintertime is detailedly discussed.Results illustrate obvious differences between normal network and abnormal network.Furthermore,this study presents the framework of proactive responses to road network vulnerability caused by negative incidents,which can be used to prevent network vulnerability,including early warning and improvements.By comparing recognition time between early warning of decoupling relationships and saturated states,results show that decoupling relationships can be recognized at least 10 minutes in advance,which provides evidence for preventing vulnerability in congestion conditions as early as possible.