Reliability Analysis Of Public Transit Operation Based On Arrival Time

Public transit operational reliability is not only one of the important index to evaluate the operating status of urban public transit,but also the key factor to consider when passengers choose the path.However,in most cities of China,because of traffic congestion during peak periods and public transit complex operating environment,which leads to bus arrival time of uncertainty,poor punctuality,bus bunch and other serious problems.In addition,increasingly fierce social and economic activities make the time of people became more and more valuable,travel time reliability of public transit became more important for passengers.Therefore,it is necessary to predict bus arrival time,which is in order to provide passengers with real-time information of public transit on operating time and reasonable travel route recommendations,and improve passengers travel time reliability.In this paper,it takes regular public transit as the research object,it proposes a dynamic forecasting model of bus arrival time,and proposes a multi-route search algorithm based on the forecast arrival time,and analyzes the travel time reliability of the bus route.Specific contents are as follows:First of all,through analysis of the impact of bus arrival time,and selecting weather,date,time,road operating time and other factors,it establishes arrival time prediction based on particle swarm optimization of wavelet neural network model.It trains and tests model by using bus survey data,results indicate that model can effectively improve the prediction speed and forecast accuracy compared with wavelet neural network model,which indicates that model has better reliability and feasibility.Secondly,it proposes a multi-route search algorithm based on the prediction of arrival time,which bases on analysis the decision of public transit passengers.The algorithm takes the transfer number of times as the constraint,and uses the forecasting model to adjust the operating time of the road section,which is in order to provide different travel route of public transit in different time period,and provide passengers with reasonable travel route advice for the same passenger.Finally,from the passenger's travel psychological intention,it proposes route operating time reliability based on the expected time.It analyzes the time reliability of travel route combined with specific examples through Monte Carlo,which tests the rationality and feasibility of method.Furthermore,the influence of three model parameters that includes risk factors of passengers,time coefficient,relative delay threshold range of passengers acceptable on the reliability of operating time is analyzed by sensitivity analysis,in order to help passengers to choose the route of travel.