Research On Passenger Travel-oriented Reliability Of A Multi-mode Rail Transit Composite Network

The rapid expansion of rail transit networks and the gradual integration of multi-mode rail transit networks have brought great convenience to passengers.Meanwhile,it brings challenges to the safe and reliable operation of rail transit networks.The core task of rail transit is to transport passengers.Therefore,research on passenger travel-oriented reliability of a Multi-mode Rail Transit Composite Network(MRCT)which consists of urban rail transit,suburban railway and high-speed railway plays an important role in improving passengers’ service quality,ensuring rail transit’s normal operations and enhancing its competitiveness.Here,the passenger travel-oriented reliability of an MRCT is studied.The expected results can enrich rail transit reliability theories and provide a certain theoretical basis and decision support for an MRCT’s reliability enhancement.The main jobs and conclusions are as follows:(1)The passengers’ travel path selection behavior and fulfillment probability of the planned trains are considered to develop a model for measuring the Connectivity Reliability(CR)of an MRCT.Passengers choose tolerable paths whose trip times are below the passengers’ tolerable trip time according to passengers’ travel path selection behavior and thus the tolerable paths are defined.A link’s CR is measured with the fulfillment probability of the planned trains running through that link;A tolerable path’s CR is measured with the product of the links’ CR that belong to the path;The CR of an OD pair is measured with the probability that there is at least one connected tolerable path between that OD pair;The average CR for each passenger traveling on a composite network is used to evaluate the CR of the composite network based on the CR evaluation for OD pairs.To evaluate the CR,a tolerable path search algorithm based on a depth-first search algorithm with branch reduction strategy,and a CR evaluation method based on a Binary Decision Diagram algorithm are proposed.(2)The model and method for measuring Time Reliability(TR)of an MRCT are proposed in terms of passengers’ whole travel process.The trip time components and their distributions are determined according to passengers’ whole travel process.Buffer time is used to measure the TR on paths;The sum of weighted TR of paths among OD pairs is used to measure TR among OD pairs in which the weights of paths are the probabilities of paths being selected by passengers;The average TR for each passenger is used to measure the TR of the composite network.A Monte Carlo simulation is applied to measure TR according to the trip time components and their distributions during the whole travel process.According to the evaluation results of TR,a multiple linear regression model is developed and solved by a stepwise regression method to quantify the relationship among TR and its influencing factors.(3)The model that considers passengers’ service quality and the probability of disturbances is developed to measure the Capacity Reliability(Ca R)of an MRCT.The Ca R of a transportation network is the probability of meeting a certain transportation demand above a certain service level.The Generalized Travel Cost(GTC)which considers crowding in trains,seat availability and the number of transfer times is computed to evaluate passengers’ service quality due to the existing literature rarely considers service level in the Ca R evaluation.The GTC increase rates on paths,residual capacity among OD pairs and a preset threshold for the GTC increase rate are used to classify categories of passengers when disturbances occur.The fraction of passengers whose travel is reliable and the passengers’ total GTC increase rate are used to measures the Ca R and capacity un-reliability of a composite network,respectively.A method for measuring the Ca R and capacity un-reliability of a composite network based on traffic assignment is proposed.(4)A model is developed for maximizing the Value of Reliability Enhancement(VRE)due to increasing train frequency on a composite network with constraints of the operational cost of increasing train frequency and the maximum train service frequencies on lines.In the model,the VRE equals the sum of the values of CR enhancement,TR enhancement and Ca R enhancement.The cost of increasing train frequency consists of the maintenance cost,train depreciation,electricity cost,and labor cost.The model is solved with a multi-population genetic algorithm.(5)To verify the effectiveness of the proposed models and algorithms,they are applied to Chengdu’s MRCT.The computation and analysis results show that: the model and method proposed here can not only evaluate the passenger travel-oriented reliability of an MRCT,but also identify critical links affecting CR,identify key factors affecting TR,as well as identify low-reliability OD pairs and lines to provide specific suggestions and guidance for improving the reliability of a composite network.The solutions for the model which maximizes the VRE due to increasing train frequency on a composite network can provide certain decision support for a rail transit company to enhance the reliability of a composite network.