Behavior And Dynamic Evacuation Guidance Strategies Of Pedestrian-vehicle Mixed Traffic On Campus

There are some special traffic characteristics on campus,such as mixed pedestrian-vehicle,paroxysmal,conflict,mutual interference,massive pedestrian and bicycle,etc.These traffic features can easily cause campus traffic congestion and traffic accidents in emergency evacuation situation.However,existing urban traffic evacuation methods mainly solve pure traffic mode evacuation,or assume travelers only use one traffic mode in a trip to select route between different transfer behavior,no consideration of the mixed traffic evacuation under the interference effects between pedestrians and vehicles.Those evacuation methods can not effectively guide pedestrian-vehicle mixed traffic flow evacuation on campus.Therefore,this dissertation focuses on the researches about the properties of campus road network,traffic flow characteristics on campus road network,campus traffic behavior model and campus traffic evacuation methods.Finally,a pedestrian-vehicle mixed traffic flow evacuation guidance scheme under campus environment is proposed.The main work is as follows:1.By analyzing the characteristics of domestic typical campus road network in colleges and universities,a division method of campus network evacuation region based on service network characteristics is proposed.In the dissertation,domestic typical campus road networks are constructed by primal approach method.Complex network theory is used to analyse the properties of campus network and prove that domestic typical campus traffic networks are small world networks,with obvious community structures.Then,based on the improved Fast-Newman algorithm,evacuation regions of campus road network are divided from the point of view of the service characteristics of campus road structure,the function of personnel activity places and evacuation exits location.And combined with network analysis method,some key nodes between campus evacuation regions are extracted,which provides the basis for the research of campus traffic behavior and traffic evacuation.2.Through the analysis of the properties of pedestrian-vehicle mixed traffic flow and traffic behavior rules on campus,the dissertation proposes the behavior model for pedestrian-vehicle mixed traffic flow on campus based on cellular automaton,which simulates the behavior of campus pedestrian-vehicle mixed traffic flow.In the dissertation,combined with courses timetable and based on statistical analysis method,the dissertation analyzes the campus pedestrian-vehicle mixed traffic flow property,spatio-temporal dynamic characteristic,interaction behavior between different traffic types,etc.And complex traffic behavior of campus pedestrian-vehicle mixed traffic flow,such as lane changing,overtaking,crossing and so on,are described.They provide research data basis for behavioral modeling and evacuation of campus traffic.To real simulate the behavior of pedestrian-vehicle mixed traffic flow on campus road network,to investigate the interaction of traffic modes from micro aspect and the influence of the interaction on traffic flow evolution from macro aspect,and to reveal complex traffic phenomena for pedestrian-vehicle mixed traffic flow,a cellular automaton model is used to describe the macroscopic and microscopic phenomena of traffic flow.According to the campus mixed traffic characteristics which are different from that of other traffic scenarios,different forward rules and lateral rules for different traffic types are proposed.Considering the influence of the lateral movement on the traffic forward speed,the proposed model increases deterministic slight deceleration before lateral movement of traffic objects based on traditional cellular automaton model.And distinguishes different acceleration for different traffic modes and same traffic mode under different traffic environment.Based on these rules,the behavior model for pedestrian-vehicle mixed traffic flow(referred to as RDPV_CA model)is proposed,which can simulate complex campus traffic phenomenon,such as running on the right side for slow traffic types,overtaking,lateral movement,interaction,synchronized flow,etc.Those simulation results are in line with the characteristics of campus complex network flow.Meanwhile,the proposed model also simulates the dynamics evolution process of campus traffic flow,the interaction among different traffic types,etc.3.To improve traffic evacuation efficiency of campus roads,based on network route load balance control theory and considered network information delay,the dissertation proposes real-time information feedback mechanism and its traffic flow route guidance strategy.To fulfill successful traffic flow route guidance for campus road network,reasonable traffic flow guidance strategies based on real-time traffic information are needed.Traditional speed,travel time,density and other network evaluation indexes can not accurately reflect campus road congestion state.Hence,they can not be used as real-time information feedback mechanism of pedestrian-vehicle mixed traffic network.In the dissertation,considering the weight of traffic object location,the weight of the road exit position and the influence of traffic type,a new real-time information feedback mechanism for pedestrian-vehicle mixed traffic flow under campus environment is proposed,named the weighted road occupancy information feedback mechanism,as the evaluation criteria of the campus traffic condition and the guidance of route selection.Firstly,for heterogeneous vehicle flow,based on the weighted road occupancy information feedback mechanism and route load balance control theory,a new traffic flow route guidance strategy for mixed traffic flow network is proposed,using NaSch model to simulate traffic dynamic evolution process under different scenarios.Then,based on the strategy,a bounded rational threshold is introduced to reduce the influence of information delay from traffic control center,and our proposed RDPV_CA model is utilized to simulate campus traffic behavior.A route guidance strategy of pedestrian-vehicle mixed traffic flow for campus network is put forward.The strategy has a better effects on route guidance,which improves the road capacity,balances traffic congestion condition on each routes,reduces the oscillations about traffic flux,velocity and traffic number,weakens the degree of interference between different traffic types.Under different scenarios,based on different cellular automaton model,the strategy shows better guidance performance for mixed traffic flow,with good robustness,applicability and adaptability.4.To improve pedestrian-vehicle mixed traffic flow evacuation efficiency of campus network,this dissertation proposes a double layer evacuation guidance scheme for campus network based on network flow control and network load balancing control theory,including dynamic traffic equilibrium assignment among regions and the route guidance strategy in regions,named evacuation guidance method based on regions division of campus road network.In the dissertation,to solve the traffic evacuation question on campus road network,evacuation optimization is carried out from the aspects of campus network as a whole and local.The double layer evacuation guidance scheme is constructed based on campus road complex network division,used to guide the evacuation of pedestrian-vehicle mixed traffic flow under campus environment.For the evacuation scheme,the traffic condition of campus network is evaluated by the weighted road occupancy.Combined with network flow control and load balancing control method,evacuation accumulation is dynamically assigned by capacity restrictive assignment method,early to avoid network congestion phenomenon and the reverse spread of congestion.The traffic flow assignment among regions and between regions is carried out by the regional model of dynamic traffic assignment which is proposed by us,combined with an improved Logit model and the boundary control strategy.And in regions,traffic flow evacuation uses real-time route guidance strategy based on the weighted road occupancy feedback mechanism.Finally,simulation of a real-world campus traffic network shows that by comparing with the dynamic user equilibrium evacuation(DUE)method,the network load and evacuation efficiency among regions tend to balance by our evacuation method,and evacuation efficiency is better than that of DUE.