Search

Archipelagos are rural regions characterized by their distinct geo-graphic layouts with many separated island communities. Water bodies hinder access to community services and make a well-functioning ferry-based public transport system the only feasible means of mass transportation. Both between islands and to connect to the mainland. Maritime environments provide great routing flexibility. However, outdated or imprecise nautical charts can hinder the safe use of such possibilities in commercial traffic. This thesis develops methods for strategic planning of archipelago transportation systems, enabling the valuation of novel fairways. Contributions are made to the fields of applied optimization and statistical modeling of spatial data considering the distinct features of archipelago environments. To improve the inference of fairways, a data fusion model for combining traditional bathymetric data and ship trajectory data from the automatic identification system (AIS) for depth inference is developed. It utilizes probabilistic modeling and survival analysis for spatial interpolation of the two qualitatively different sources of information. To handle the large amounts of automatic identification system (AIS) data and the non-stationarity cause by holes (islands) in the model region, a discretized mesh approach is used. In the second step an optimization model for joint network design and scheduling of ferry public transport, with the capability of evaluating currently unavailable fairways with opening costs. In this study, characteristic features of archipelago ferry networks are identified. These features are then expressed using a modified vehicle routing model with multiple trips. Results show that factors such as traffic requirements, operating costs and fleet composition has a large impact on the value of a fairway. In summary, the thesis develops methods for inferring novel fairways using the alternate source of information provided by ship trajectory data., Skärgårdar utgör landsbygdsområden kännetecknade av sina särskilda geografiska förutsättningar. Vatten separerar ösamhällen och utgör fysiska begränsningar som påverkar tillgången till samhällstjänster. De gör ett väl fungerande vattenburet kollektivtrafiksystem centralt för att möjliggöra god tillgänglighet både mellan öarna och till fastlandet för alla invånare. Marina miljöer erbjuder vidare stor flexibilitet i ruttplaneringen. Däremot kan föråldrade eller oprecisa sjökort hindra säker användning av sådana möjligheter i kommersiell trafik. Denna avhandling utvecklar metoder för att förbättra den strategiska planering av skärgårdars transportsystem och möjliggöra värdering av nya farleder. Bidrag görs inom tillämpad optimering och rumslig statistisk modellering som tar hänsyn till de särskilda egenskaperna skärgårdsmiljöer besitter. För att förbättra skattningen av farleder utvecklas en datafusionsmodell som kombinerar traditionella djupdata med fartygstrajektorier från Automatic Identification System (AIS). Probabilistisk modellering och överlevnadsanalys används för rumslig interpolering av de två kvalitativt olika informationskällorna. För att hantera de stora datamängderna i AIS-datat och icke-stationäriteten som introduceras av hål (öar) i studieområdet används en diskretiserad modelleringsansats. I det andra steget utvecklas en optimeringsmodell för kombinerad nätverksdesign och schemaläggning av färjetrafiken. Modellen, som baseras på en modifierad ruttplaneringsmodell med flera resor, inkluderar många av de karakteristiska drag för skärgårdstrafik som identifierats och erbjuder möjlighet att utvärdera för närvarande otillgängliga korridorer med fasta öppningskostnader. Resultaten visar att faktorer såsom trafikeringskrav, driftskostnader och fordonsflottans sammansättning bör tas i beaktande vid värdering av nya farleder. Sammanfattningsvis pekar avhandlingen på hur förbättrad djupinformation kan hjälpa till att härleda möjliga transportkorridorer som inte, Funding: Financing was provided by the Swedish Transport Administration (Trafikverket) via the Swedish Maritime Administration (Sjöfartsverket) under grant TRV 2019/119584.

In this paper we propose a model for joint ferry network design and scheduling in archipelago public transport systems. The model uses a multi-trip vehicle routing approach with inbound-outbound and returning trip structures where traffic requirements are implemented as window requirements. A case study is performed scheduling a full day of traffic in the southern archipelago outside of Gothenburg, Sweden. Results indicate that this approach can capture the many different stopping patterns of a line seen in real systems and produce realistic time-tables. Computational performance is shown to be highly dependent on the number of window constraints included., Funding Agencies|Swedish Transport Administration through project OSKAR [TRV 2019/119584]

Traffic state estimation is an important part of the traffic control process and aims to creates an accurate understanding of the current situation in traffic system. Bayesian Filtering is a statistical modelling framework that is useful in representing traffic state update as well as the relation between traffic state and detection data. This study develops a hybrid approach and uses non-parametric Gaussian Process (GP) to model the state-space transition of traffic system. Through representing the system models as either fully data-driven GP or as a hybrid model using a parametric mean function fusing the conventional principle of traffic flow with the data-driven approach, the requirement of an analytical model can be removed or relaxed. The computational results show that the proposed approach for lane based TSE can capture both short-term fluctuations and larger demand changes. In particular, the Bayesian nature of the GP models offer relative ease in quantifying the model uncertainties in combination with a conventional traffic flow model., QC 20220929Part of proceedings: ISBN 978-1-7281-9142-3

Traffic signals play an important role in a citys traffic system both as an instrument fordistribution of space and for creating safety. When planning traffic signals there are a numberof different forms of control that can be used. One of there is coordinated control where anumber of traffic signals are planned in union in order to create a green wave along an arterialroad.The green wave works under the assumption that the traffic travelling on the link is keepingthe appropriate speed. Therefore there is also a demand that the planner has the correctinformation about the speed distribution of the vehicle composition using the system.In Stockholm there is a political will to prioritize walking and cycling as modes of transportahead of private transport by car in the planning. This is expressed in the official documentexplaining the strategies for mobility (Sv. “Framkomlighetsstrategin”). An important tool inmanaging to make the change in the traffic system that this would entail is traffic signaltiming. The green waves that have been planned historically have been planned for theautomobile for the speeds that are relevant for this mode of transport without necessarilypaying much attention to the effects this would have on cyclists.Götgatan is a street in central Stockholm that is an important connection for the cyclistscommuting from the southern suburbs to the central part of the city. In 2014 the municipalitydecided to rebuild this street to widen the bike lanes and implement a green wave designed tobe appropriate for a cyclist travelling at a speed of roughly 18 km/h. There was no hard datato support this assumed speed and therefore the efficacy of the wave is also uncertain for theactual cyclists using the street. Therefore this project has been directed at measuring the actualdistribution of velocities found in the afternoon rush. This was done for normal cyclists andelectrically assisted cyclists to try to quantify the effect of an increased share of electrica