Automated timetable design for demand-oriented service on suburban railways

Abstract

Economic and attractive operation of suburban railways can only be realised by flexibilisation of headways, adaptation of the network and capacity of the different lines. Under those circumstances, the computation of optimal operation programmes is very complex. This contribution presents a two-level approach (computation of transport offer, timetable design) and shows the results obtained from fullyautomatic offer planning and timetabling for a suburban railway. MotivationDemand-oriented metro systems have proven their commercial success for more than 20 years in practice (Strobel and Schütte 2004). The reason of success for this kind of transport system is the so-called flexible operation: the headways between the small capacity trains are chosen in such a way (temporal flexibilisation) that the trains are equally high occupied. One example of this operating strategy is the fully automatic underground metro (AGT-Automated Guided Transit) of the city of Lille in northern France, which operates at headways between 1 min and 6 min (cf. Fig. 1).Suburban railway systems in medium-sized conurbations, on the contrary, are mainly operated with long trains at long constant headways most of the day. This paper examines how the flexible operation, i.e., using small vehicles at short intervals for the strict adaptation of transport offer to demand, can be introduced on suburban railway systems and how a timetable can be created in this case.The literature on demand-oriented timetabling of public transit systems mainly focusses on two points:

The optimization of train timetable stop-skipping patterns in urban railway operations

“…Shafia et al [3] proposed a robust timetabling model, and proposed a robustness measure to compute the required buffer intervals. Beside mainline train timetabling problem, in urban transport system, Albrecht [4] proposed a two level approach to generate demand-oriented timetable, where the optimal train frequency and the capacity of trains are first determined and then the schedule of trains are optimized. Wang et al [5] proposed a detailed non-linear bi-level model of train movements with stop-skipping and the O-D dependent passenger demands, as well as a genetic alg.…”

“…This parameter affects the right-hand side of inequality (4). As the objective function assumed to be crisp, the Werner's approach is applicable.…”

Passenger-demands-oriented train scheduling for an urban rail transit network

“…The algorithm proposed by Assis and Milani (2004) can effectively generate train schedules for the whole day. Furthermore, a demand-oriented timetable design has been proposed by Albrecht (2009), where the optimal train frequency and the capacity of trains are first determined and then the schedules of the trains are optimized. Niu and Zhou (2013) optimized the train schedules for an urban rail transit line with consideration of time-varying origindestination passenger demands in heavily congested situations.…”

Some Aspects of Heuristic Algorithms and Their Application in Decision Support Tools for Freight Railway Traffic Organization

“…The proper preparation of a train timetable plays a major role in this process, due to the specificity of rail traffic, that is, to place the train path in time and space in a safe manner, respecting the needs of the customers. The problem with timetable construction involves the determination of train paths based on the traffic graph [1,4,8,9,17,21]. In the process of assigning routes for trains, the priorities on the traffic graph should be taken into account [20], as well as multiple technical and operational factors, including railway traffic regulations [19].…”