Prediction Model On The Recovery Time Of Traffic System With Impactive Demand Disturbance

With rapid social and economic development, urban transport has been witnessed unprecedentedchanges, which has led to increasingly prominent traffic problems such as intense contradictionbetween supply and demand. Now, the transport sector is facing increasing pressures and challenges.The study of demand impact on the transportation system is not only the basis of passenger transportdevelopment strategy and macro-control method, but also the premise of the emergency trafficmanagement. In this paper, we suppose numbers of passengers are stranded in certain transport sites,then we study the effect of demand impact under circumstances of government’s provision ofinformation support or not and propose forecasting model on recovery time of traffic sites bytheoretical analysis and assumptions. The specific works are as follows:Firstly, we summarize the conceptions, types and features of demand impact based on the resultsof existing research, and further analyze the demand impact of passenger in traffic site.Secondly, under circumstances of government’s provision of information support or not,maximum-entropy model with full information and cellular automata model with incompleteinformation are respectively proposed.Based on government’s provision of information support, the pressure distribution of alltransport sites are completely known by the stranded passengers. With complete information, everypassenger would reach the maximum free space according to utility maximization. Based on utilitymaximization and maximum-entropy principle, we establish maximum-entropy model of transferprobability by which recovery time of all sites can be gotten in large-scale network.In the absence of information support, the stranded passengers could present irrationally due touncertainty of information. According to the differences in conformity preferences, the strandedpassengers are divided into three types: independent, environment-adaptive and conformity. Then, wedesign state sets and evolution rules for different types cells to build the cellular automata simulationmodel of recovery time, which is convenient to be applied into observation of individual behaviordecisions’ impact on recovery time.Finally, after case studies, we find information guidance help reduce the pressure of sites andshorten recovery time: transparent information make passengers’ decision-making process shorter andinduce more of them to give up trips; also, it can decrease the blind transfer of passengers, which isapt to cause system oscillation and unnecessary security risks.