Travel Services Design And Reliability Evaluation

The expansion of China’s railway passenger network, the popularization of online ticketing and growing passenger travel demand have laid the foundation for the development of travel services. Researches on travel services aim to provide travelers with personalized travel plans that cover their whole journey. The researches have a significant influence to improve the service quality of travel service providers and agencies. This paper proposes a generalized travel services problem and dissects it into two subproblems, design and evaluation. From the perspective of railway travelers, this paper focuses on the model and algorithms for railway travel planning problem, as well as methods on reliability evaluation of railway travel plans. The main research work of this paper can be concluded as follows:1. For all intents and purposes, a traveler’s journey consists of a series of activities, so the whole journey can be abstracted as a travel plan consisted of travel services that support travel activities. In order to design travel plans that cover the whole journey the paper proposes a novel problem-travel services design problem (TSDP) from the perspective of combining time and space. A generalized model and solution of TSDP is proposed, which breaks the limitation that existent researches separate time and space problems.2. From the perspective of evaluating the reliability of travel plans, the paper presents travel services evaluation problem (TSEP). Common characteristics of TSEP in different fields are analyzed, and the paper theoretically proposes the general evaluation methods on reliability of travel time and service capacity.3. For railway travel planning problem, a transit service network is constructed to describe railway passenger services. The definition of railway travel plan is given. A railway travel planning model with personalized optimization objectives is proposed, and it models a realistic travel planning problem with non-designated transfer points and possible co-station transfers.4. To solve railway travel planning model, the paper proposes two algorithms with respect to different application scope. First algorithm is called transfer node matching. Following a two-stage strategy, it searches a set of feasible travel plans within a certain number of transfers and then from the set selects optimal travel plans. The other algorithm is called ranking-based travel plan search. When searching travel plans, it chooses alternative transit cities from a selective set of prioritized cities. As it saves search space, the algorithm can efficiently find feasible travel plans. Examples of both algorithms are given.5. For railway travel plan reliability evaluation, it discusses the levels of reliability researches and factors that affect travel time reliability and service capacity reliability. From the perspective of service reliability, the paper focuses on the following two issues: travel time reliability evaluation and service capacity reliability evaluation.6. In travel time reliability evaluation, methods to measure travel time reliability of a single-legged travel plan and connecting reliability of a multi-legged travel plan are proposed. Travel time reliability and connecting reliability can be calculated based on trains’ punctuality data.7. In service capacity reliability evaluation, considering the status quo of China Railway Ticketing System’s data structure, the paper proposes a method to assess the train’s service capacity reliability by measuring the degree of difficulty for passengers to buy tickets. The probability of tickets in short supply is calculated by analyzing the process of passengers buying tickets, and then the service capacity reliability of a travel plan is measured based on the probabilities.