Research On Realization Of Optimal Travel Time Route For Day-to-day Dynamics

In order to satisfy the travelers’ demand for avoiding congestion,optimizing travel route and minimizing travel time,this paper explores the way to realize the optimal route of travel time from two aspects:one is exploring the regularity of road traffic flow of peak period;another is the day-to-day route choice game behavior evolution process.By analyzing and processing the RFID(radio frequency identification)data of the downtown area of Nanjing City,this paper studies the characteristics of day-to-day route choices in the reallife road network.It is found that in the vast majority of the cases,the actual travel route of a traveler is neither the shortest travel distance route nor the shortest travel time route,but closer to the shortest travel distance route.This rule is particularly prominent in the morning and evening peak hours.Based on the statistics of road traffic flow from RFID data,it is found that the actual stochastic time-varying characteristics of road traffic flow are that the travel time of a road segment changes randomly and is independent of time,and the values approximately obey the normal distribution statistically.A high probability algorithm for realizing the shortest travel time route is proposed for individual day-to-day travellers.The counterexample demonstrates that the existing adaptive algorithms are ineffective in solving the shortest travel time route problem with the characteristics of travel time of a road segment changes randomly and is independent of time.Furthermore,based on the historical probability distribution of travel time,the traveler’s risk measurement preference and the current road traffic flow situation,a shortest travel time route high probability realization algorithm named HPS for day-to-day traveller is proposed,which can ensure that the shortest travel time route be realized with high probability.The realization of traffic equilibrium with optimal travel time is studied for the overall of day-to-day travellers.Based on the analysis of the equivalence between traffic equilibrium and the optimal travel route of all travelers in the road network,by analyzing the route choices in the reallife road network,it is found that the realroad traffic flow,which is the traffic flow under the condition of non-public information as well,is unbalanced.In view of the existing research especially the disputes about whether the traffic flow can reach equilibrium under the condition of public information,a human-computer interactive experiment simulation system and a multi-agent simulation system are established on the basis of simple road network for studying the day-to-day route choice evolution process.Experiments and simulation results show that it is difficult to realize equilibrium of road traffic flow under the condition of public information,which makes it impossible to realize the optimal travel time for all travelers,and the reasons for the disputes of equilibrium are explained as well.The stochastic nonequilibrium user optimal(SNEUO)state is defined,and the causes of non-SNEUO state are analyzed.The data analysis shows that the travelers’ actual travel times on each alternative route can be approximated to the optimal state under the condition of the alternative routes being undifferentiated,while the actual travel times generally deviates from the optimal state under the condition of the alternative routes being differentiated.The state of expected value of travel times on each alternative routes during the whole day-to-day route choice evolution process approximately equals to Wardrop equilibrium value under the condition of the alternative routes being undifferentiated,is defined as the stochastic non-equilibrium user optimal state(SNEUO),and developed into the generalized SNEUO state under the condition of the alternative routes being differentiated.The analysis shows that the actual travel times on both alternative routes under the condition of the alternative routes being differentiated are not the SNEUO states.Furthermore,it is demonstrated that travel distance is the reason that affects the traveler’s route choice,which leads to the traveler’s preference to choose shorter route,thus forming the non-SNEUO state under the condition of the routes being differentiated.The possibility of realizing SNEUO state under the condition of the alternative routes being differentiated is studied.The factors that influence and feasible to adjust the travelers’ route choices are analyzed.Based on the route choice mechanism,the principles of two feasible strategies,traffic information and congestion pricing,are analyzed to adjust the traveler’s route choice.Furthermore,based on human-computer interactive experiments,the possibility of realizing SNEUO state under differentconditions of traffic information disclosure is verified.Based on multi-agent simulation,the possibility of realizing SNEUO state by congestion charging and the corresponding value of charging are determined.Human-computer interactive experiments show that the SNEUO status can not be achieved by disclosing information of route choice results of last round and recent historical statistics of the route choice results to different proportions of travelers,but it is more advantageous to disclose recent historical statistics to tend to SNEUO status.The SNEUO state can be achieved by disclosing to the proper proportion(about 65%)of travelers with the statistical mean values of travel times in non-SNEUO state.The simulation results of multi-agent show that there is a reasonable congestion value of charging to realize the SNEUO state under the condition of the routes being differentiated.