Simulation Of Network Traffic Flow Evolution Under Different Information Conditions

The services of Advanced Traveler Information Systems(ATIS)are progressed with the development of information technology,whose width(the scope of servicing population)and depth(the precision of traffic information)are upgraded gradually.However,the evolution process of road network traffic flow changes as well whilst more information intervening the traffic system.Based on analyzing three main traffic system elements,which are traffic information,travelers and evolution of network traffic flow,the systematic dynamic evolution mechanism among them is studied in this paper.Focusing on revealing the rules and the features of the network traffic flow evolution,and on the impact of travelers' choosing the route incurring from traffic information,the paper discusses the influences to network traffic flow evolution caused by modes,targets and contents of the information feedback.After building the traffic network,the model of network traffic flow evolution affected by information is formulated,combining with the principle of the day-to-day evolution.Moreover,the route choice procedure of heterogeneous travelers is described by the disaggregate model(such as the Logit model).Meanwhile,the emulation system of network traffic flow evolution with the information feedback scenario is designed according to the principle of traffic system simulation.The multi-agent method is introduced to imitate all real travelers and the Monte-Carlo method is used to simulate the travelers' route choice process.Regarding the information feedback modes,there are four approaches such as providing travelers with the traveling time of the most expedite route,the most congested route and all routes.Each information feedback mode can direct the relevant special network traffic flow evolution process.Providing all travelers with traffic information may not essentially decrease the overall average traveling time,whereas,unsuitable information feedback method may extend that time and lead to the traffic flow vibration.In comparison,providing travelers with the travel time of the most congested route will definitely make sense.From the perspective of the information feedback targets,the number of travelers is large and their features are various,which cause the obvious heterogeneous characteristics,that is,each individual could choose different route due to his special cognition even if travelling in the very same traffic environment.Therefore,the paper researches the following five aspects to explain the effects of travelers' heterogeneity on the network traffic flow evolution process,including the degree of travelers' rationality degree heterogeneity,the route habitual factor,the individual choice preference,the information compliance and the value of time.There is close relationship between the network traffic flow evolution process and the features of travelers' heterogeneity.Hence,the less the diversity degree among travelers and the more their rational levels are,the clearer the network traffic flow aggregated phenomena is and the stronger the vibration is,the travelers' mean traveling time keeps longer as well.On the contrary,the distribution of the traffic flow is more even,the vibration is gentler and the mean time is shorter.Finally,corresponding to the information feedback content,a systematic framework of individual route recommendation is implemented to meet all travelers' personalized requirements.In this framework,four main traveling preference elements(time,grade,toll and distance)are selected to formulate a route recommendation model.Additionally,the optimal route is recommended by the similarity between the vector of individual preferences and the vector of the route's features,according to the Pearson similar coefficient.There is clear correlation among the features of both the network traffic flow evolution and the travelers' preferences,which is affected dramatically by the traveling demands.During the evolutional process,the traffic flow can reach equilibrium states or vibrating equilibrium ones,and the individual information compliance can also get acceptable level.Apart from meeting those individual requirements,this optimal route recommendation method balances the distribution of network traffic flow.Taking the congestion charging in the traffic network as the background,the congestion pricing strategy adapting to the network traffic flow evolution is researched.By this charging method,the travelers' route choices are optimized,the distribution of network traffic flow is balanced and the personal average traveling time becomes shorter simultaneously.