Crespo Materna, Arturo (2020)
Dynamic and Intermodal Disruption-Management for Commuter Railway Networks.
Technische Universität Darmstadt
doi: 10.25534/tuprints-00011662
Dissertation, Erstveröffentlichung
Kurzbeschreibung (Abstract)
Railway systems are particularly vulnerable to the occurrence of unexpected events and disruptions due to their size and the complex arrangement of their different components and operations (i.e. stations, tracks, switches, vehicles, personnel, passenger traffic, signals, operation control, schedules, etc.). As disruptions for any critical infrastructure are inevitable, decision-makers need to establish strategies aimed at guaranteeing the operational continuity of the systems and upholding basic service qualities during these events. In many railway networks, planned disruption-management approaches have been established for a structured reaction to the disruption. Planned disruption-management approaches foresee the development and implementation of disruption programs (DRPs). DRPs can be explained as sets of pre-defined dispatching measures for trains, rerouting measures for disrupted passengers and communication protocols for staff members, which are developed to address specific disruption scenarios within a railway network. In consequence, every DRP contains a set of line-specific measures (i.e. valid for a whole line), which is mainly constituted by two different concepts, namely, an operating concept and a transport concept. The operating concept contains a series of line-specific measures that allow the operating program of a line to adjust to the degraded infrastructure availability induced by the disruption. The transport concept contains a series of passenger rerouting measures that address the affected serviceability of the system. DRPs are implemented across different phases. Among these phases, the transition phase to stable operations is arguably the most critical. The transition phase lasts from the declaration of the chosen DRP and until the system has reached stable operations during the disruption. Stability is only achieved when all trains run reliably on their DRP envisioned (shortened) routes with (reduced) frequencies, without accumulating delay. To date, dispatchers execute the implementation of DRPs manually. Therefore, the successful deployment of a DRP on a disrupted network is still critically influenced by the experience and skill of highly strained dispatchers. DRP development itself is also a strenuous process, which requires the involvement of different stakeholders (e.g. experienced dispatchers, public transport operators and others). While some work has been aimed at transitioning from a manual development of DRPs to a development assisted by a decision-support system, there are still gaps to be filled. The evaluation of intermodal passenger rerouting measures within transport concepts requires particular attention. The work presented throughout this document has two specific objectives. The first objective is to develop a model that allows decision-makers to develop passenger intermodal rerouting strategies for DRP transport concepts, which consider the residual capacity of local public transport systems. The second objective is the development of a system capable of supporting the dynamic deployment of DRP operating concepts on an actual operating situation of the network. Each of the objectives focuses on the different concepts within DRPs, yet they are unrelated to one another. Therefore, this work is divided into two different Sections. The first Section describes the development of a model for estimating the residual capacity of the public transport means utilized for the intermodal rerouting of passengers, as foreseen in the DRP transport concept under investigation. The assessed passenger rerouting strategies provide a foundation for a subsequent and much more in-depth discussions with local public transport operators. To support an estimation of the residual capacity, an assessment framework is derived and validated utilizing actual operational information from three public transport modes namely, buses, light rail and subway. The model incorporates the assessment framework and is structured using rule-based algorithms supported by graph theory. The model has the capability to incorporate different rerouting strategies from DRP transport concepts. Furthermore, the model is designed for general validity and be applicable regardless of the implementation environment (i.e. means of public transport and layout of the public transport and disrupted railway networks). The second Section constitutes the main contribution of this work. It introduces a system for the dynamic deployment of the line-specific DRP operating concept to the actual operating situation of a disrupted commuter railway network. The actual disrupted situation considers the time of day, the actual position of the trains in the network, and the affected infrastructural elements. As a result, the system delivers a train number and minute-specific (i.e. with an accuracy of seconds) conflict-free schedule that ensures the network is able to reach stable operations. The dynamic DRP deployment system is conceived in a modular structure so that its modules can be easily updated or replaced and eventually, new modules can be added. Each one of the modules that constitute the system is derived and discussed in detail throughout this work. In line with the problem, the dynamic DRP deployment system foresees the adjustment of both the schedule and circulation plans of the disrupted commuter railway network. The system is structured through heuristic (i.e. heuristic conflict identification - or CD - and conflict resolution - or CR - approaches) and metaheuristic methods (i.e. Genetic and Tabu Search algorithms) to address the mostly NP-hard problems. Additionally, the system divides the overall problem into two different operational levels: line-specific and vehicle-specific. At the line-specific level, the approach identifies and classifies line-specific conflicts (including vehicle availably and reachability conflicts) and establishes potential conflict solution alternatives with support of the DRP operating concept as well as a predefined set of eight types elemental conflict solutions. The potential solution alternatives are implemented on individual trains in order to solve the line-specific conflicts and establish a set of conflict resolution alternatives at the line-specific operational level. The alternatives that have been generated at the line-specific operational level are combined through metaheuristic algorithms. Each combination is later handled at the vehicle-specific operational level. A heuristic vehicle-specific CDCR approach identifies and resolves any induced conflict to obtain a conflict-free schedule from every combination. At this level, the CDCR process handles four different types of conflicts. These include occupancy, infrastructure availability, circulation and service conflicts. For every identified conflict, the CDCR process develops potential conflict resolution alternatives utilizing six types of elemental conflict solution alternatives. The conflict resolution alternatives are evaluated by contemplating aspects such as the expected relative-time change, the induced change on the projected operating situation, changes on the platform tracks and train service or train stop cancellations. Once assessed, only one conflict resolution alternative is selected and implemented. Furthermore, once the combinations are conflict-free, they are evaluated according to already assessed information from every conflict resolution utilized to derive the conflict-free combination. This is complemented by an evaluation of the induced end-of-day imbalances (i.e. vehicles that terminate their duties at locations in the network that are not compatible with the scheduled operations on the next day) and changes in the turning stations of the trains. Finally, having ascertained the fitness of every conflict-free combination, the system is able to select and display the best alternative. Ultimately, the contribution advanced in the second Section of this work lays the foundation for a semi-automated deployment of the DRP operating concepts in the actual operating situation of the network. The dynamic DRP deployment system is strictly designed to uphold the prompt transition of the network toward stable operations. While the system is designed for commuter railway networks, it can be employed regardless of their local implementation environment.
Typ des Eintrags: | Dissertation | ||||||
---|---|---|---|---|---|---|---|
Erschienen: | 2020 | ||||||
Autor(en): | Crespo Materna, Arturo | ||||||
Art des Eintrags: | Erstveröffentlichung | ||||||
Titel: | Dynamic and Intermodal Disruption-Management for Commuter Railway Networks | ||||||
Sprache: | Englisch | ||||||
Referenten: | Oetting, Prof. Dr. Andreas ; Martin, Prof. Dr. Ullrich | ||||||
Publikationsjahr: | April 2020 | ||||||
Ort: | Darmstadt | ||||||
Reihe: | Schriftenreihe der Institute für Verkehr | ||||||
Band einer Reihe: | 12 | ||||||
DOI: | 10.25534/tuprints-00011662 | ||||||
URL / URN: | https://tuprints.ulb.tu-darmstadt.de/11662 | ||||||
Kurzbeschreibung (Abstract): | Railway systems are particularly vulnerable to the occurrence of unexpected events and disruptions due to their size and the complex arrangement of their different components and operations (i.e. stations, tracks, switches, vehicles, personnel, passenger traffic, signals, operation control, schedules, etc.). As disruptions for any critical infrastructure are inevitable, decision-makers need to establish strategies aimed at guaranteeing the operational continuity of the systems and upholding basic service qualities during these events. In many railway networks, planned disruption-management approaches have been established for a structured reaction to the disruption. Planned disruption-management approaches foresee the development and implementation of disruption programs (DRPs). DRPs can be explained as sets of pre-defined dispatching measures for trains, rerouting measures for disrupted passengers and communication protocols for staff members, which are developed to address specific disruption scenarios within a railway network. In consequence, every DRP contains a set of line-specific measures (i.e. valid for a whole line), which is mainly constituted by two different concepts, namely, an operating concept and a transport concept. The operating concept contains a series of line-specific measures that allow the operating program of a line to adjust to the degraded infrastructure availability induced by the disruption. The transport concept contains a series of passenger rerouting measures that address the affected serviceability of the system. DRPs are implemented across different phases. Among these phases, the transition phase to stable operations is arguably the most critical. The transition phase lasts from the declaration of the chosen DRP and until the system has reached stable operations during the disruption. Stability is only achieved when all trains run reliably on their DRP envisioned (shortened) routes with (reduced) frequencies, without accumulating delay. To date, dispatchers execute the implementation of DRPs manually. Therefore, the successful deployment of a DRP on a disrupted network is still critically influenced by the experience and skill of highly strained dispatchers. DRP development itself is also a strenuous process, which requires the involvement of different stakeholders (e.g. experienced dispatchers, public transport operators and others). While some work has been aimed at transitioning from a manual development of DRPs to a development assisted by a decision-support system, there are still gaps to be filled. The evaluation of intermodal passenger rerouting measures within transport concepts requires particular attention. The work presented throughout this document has two specific objectives. The first objective is to develop a model that allows decision-makers to develop passenger intermodal rerouting strategies for DRP transport concepts, which consider the residual capacity of local public transport systems. The second objective is the development of a system capable of supporting the dynamic deployment of DRP operating concepts on an actual operating situation of the network. Each of the objectives focuses on the different concepts within DRPs, yet they are unrelated to one another. Therefore, this work is divided into two different Sections. The first Section describes the development of a model for estimating the residual capacity of the public transport means utilized for the intermodal rerouting of passengers, as foreseen in the DRP transport concept under investigation. The assessed passenger rerouting strategies provide a foundation for a subsequent and much more in-depth discussions with local public transport operators. To support an estimation of the residual capacity, an assessment framework is derived and validated utilizing actual operational information from three public transport modes namely, buses, light rail and subway. The model incorporates the assessment framework and is structured using rule-based algorithms supported by graph theory. The model has the capability to incorporate different rerouting strategies from DRP transport concepts. Furthermore, the model is designed for general validity and be applicable regardless of the implementation environment (i.e. means of public transport and layout of the public transport and disrupted railway networks). The second Section constitutes the main contribution of this work. It introduces a system for the dynamic deployment of the line-specific DRP operating concept to the actual operating situation of a disrupted commuter railway network. The actual disrupted situation considers the time of day, the actual position of the trains in the network, and the affected infrastructural elements. As a result, the system delivers a train number and minute-specific (i.e. with an accuracy of seconds) conflict-free schedule that ensures the network is able to reach stable operations. The dynamic DRP deployment system is conceived in a modular structure so that its modules can be easily updated or replaced and eventually, new modules can be added. Each one of the modules that constitute the system is derived and discussed in detail throughout this work. In line with the problem, the dynamic DRP deployment system foresees the adjustment of both the schedule and circulation plans of the disrupted commuter railway network. The system is structured through heuristic (i.e. heuristic conflict identification - or CD - and conflict resolution - or CR - approaches) and metaheuristic methods (i.e. Genetic and Tabu Search algorithms) to address the mostly NP-hard problems. Additionally, the system divides the overall problem into two different operational levels: line-specific and vehicle-specific. At the line-specific level, the approach identifies and classifies line-specific conflicts (including vehicle availably and reachability conflicts) and establishes potential conflict solution alternatives with support of the DRP operating concept as well as a predefined set of eight types elemental conflict solutions. The potential solution alternatives are implemented on individual trains in order to solve the line-specific conflicts and establish a set of conflict resolution alternatives at the line-specific operational level. The alternatives that have been generated at the line-specific operational level are combined through metaheuristic algorithms. Each combination is later handled at the vehicle-specific operational level. A heuristic vehicle-specific CDCR approach identifies and resolves any induced conflict to obtain a conflict-free schedule from every combination. At this level, the CDCR process handles four different types of conflicts. These include occupancy, infrastructure availability, circulation and service conflicts. For every identified conflict, the CDCR process develops potential conflict resolution alternatives utilizing six types of elemental conflict solution alternatives. The conflict resolution alternatives are evaluated by contemplating aspects such as the expected relative-time change, the induced change on the projected operating situation, changes on the platform tracks and train service or train stop cancellations. Once assessed, only one conflict resolution alternative is selected and implemented. Furthermore, once the combinations are conflict-free, they are evaluated according to already assessed information from every conflict resolution utilized to derive the conflict-free combination. This is complemented by an evaluation of the induced end-of-day imbalances (i.e. vehicles that terminate their duties at locations in the network that are not compatible with the scheduled operations on the next day) and changes in the turning stations of the trains. Finally, having ascertained the fitness of every conflict-free combination, the system is able to select and display the best alternative. Ultimately, the contribution advanced in the second Section of this work lays the foundation for a semi-automated deployment of the DRP operating concepts in the actual operating situation of the network. The dynamic DRP deployment system is strictly designed to uphold the prompt transition of the network toward stable operations. While the system is designed for commuter railway networks, it can be employed regardless of their local implementation environment. |
||||||
Alternatives oder übersetztes Abstract: |
|
||||||
URN: | urn:nbn:de:tuda-tuprints-116626 | ||||||
Sachgruppe der Dewey Dezimalklassifikatin (DDC): | 600 Technik, Medizin, angewandte Wissenschaften > 620 Ingenieurwissenschaften und Maschinenbau | ||||||
Fachbereich(e)/-gebiet(e): | 13 Fachbereich Bau- und Umweltingenieurwissenschaften 13 Fachbereich Bau- und Umweltingenieurwissenschaften > Verbund Institute für Verkehr 13 Fachbereich Bau- und Umweltingenieurwissenschaften > Verbund Institute für Verkehr > Institut für Bahnsysteme und Bahntechnik |
||||||
Hinterlegungsdatum: | 02 Jun 2020 11:57 | ||||||
Letzte Änderung: | 27 Nov 2020 10:17 | ||||||
PPN: | |||||||
Referenten: | Oetting, Prof. Dr. Andreas ; Martin, Prof. Dr. Ullrich | ||||||
Export: | |||||||
Suche nach Titel in: | TUfind oder in Google |
Frage zum Eintrag |
Optionen (nur für Redakteure)
Redaktionelle Details anzeigen |