Coordinated planning of inland vessels for large seaports

Shijie Li

Research output: ThesisDissertation (TU Delft)

47 Downloads (Pure)

Abstract

Seaports are crucial nodes in international trade and transport. Some of the cargoes arriving at seaports are transshipped to other ports, while others are transported to inland destinations. Every time an inland container vessel enters the port, it calls at many different terminals spread over the port area. Two coordination problems exist in the planning of inland vessels in large seaports: firstly, the long stay in the port and secondly, the insufficient terminal and quay planning with respect to the sailing schedules of sea-going vessels and inland vessels. To solve these problems, this thesis aims to improve the reliability and efficiency of inland vessel transport in seaports. To achieve this, efficient handling of inland container vessels in large seaports is required. This could improve the efficiency and reliability of inland waterway transport from seaports to hinterland and vice versa. Meanwhile, this could also contribute to enhancing of the inter-terminal transport in large seaports, as a potential solution for alleviating congestion on roads. Moreover, efficient handling of inland vessels could facilitate flexible planning of transport over water, so that this transport mode can be better integrated into the synchromodal transport chain.
Therefore, three classes of automatic coordination methods are proposed in this thesis for seaports with different sizes: for small seaports, a partially-cooperative coordination method with single-level interaction based on distributed constraint optimization is proposed; for medium-sized seaports, a partially-cooperative coordination method with multi-level interactions based on MIP, coordination rules and constraint programming, is proposed; for large seaports, a fully-cooperative coordination method with multi-level interactions based on Benders decomposition and large neighborhood search is proposed.
With the proposed methods, firstly, the vessel operators can decide to what extent they would like to be coordinated, either partially-cooperative or fully-cooperative; secondly, terminal operators can also use these methods to estimate how much time each inland vessels spends at the terminal, in order to determine the schedules of terminal operations; thirdly, whenever real-time disturbances or accidents happen, the previously generated rotations might be no longer optimal, the proposed methods can take into account these disturbances and generate new and better rotations for vessel operators based on the up-to-date information. Simulation results demonstrate that the proposed approach can significantly reduce the round-trip time the inland vessels spend in the port, as well as reducing the time they spend waiting at container terminals.
Moreover, from the perspective of using inland vessels also for inter-terminal transport (ITT) in seaports, vessel operators can use the proposed methods to decide whether they are willing to cooperate to take extra ITT containers based on the possible extra round-trip time and waiting time that they may spend. Moreover, terminal operators can also estimate how much extra ITT containers could be transported by the incoming vessels during different times of a day, so that they can plan the ITT containers to be transported by inland vessels accordingly.
To conclude, this thesis investigates the operational planning of inland vessels in seaports. This thesis shows the potential of the proposed new approaches for improving the efficiency and reliability of inland vessel transport in seaports.
Original languageEnglish
Awarding Institution
  • Delft University of Technology
Supervisors/Advisors
  • Lodewijks, Gabriel, Supervisor
  • Negenborn, R.R., Advisor
Award date13 Dec 2016
Print ISBNs978-90-5584-216-2
DOIs
Publication statusPublished - 2016

Fingerprint

Dive into the research topics of 'Coordinated planning of inland vessels for large seaports'. Together they form a unique fingerprint.

Cite this