General Information
This page comprises supplementary data analysis results and
datasets for container ship traffic models.
The datasets are built on the basis of probabilistic models derived
from real container ship traffic data of five ports:
Hamburg,
Le Havre,
Rotterdam,
Shanghai,
Singapore
(by the rules of the granted access we are not allowed to release
the source data.)
Values of the following parameters were drawn pseudo-randomly
according to the probabilistic models:
• number of the ship arrivals n,
• ship ready times, r_j,
• ship processing times, p_j,
• ship lengths, L_j,
• ship importance (or value, weight), w_j.
The generated datasets retain features of the original traffic such as:
- ship size distributions,
- ship processing time distributions and their dependence on the ship sizes,
- ship return interval distributions and their dependence on the ship sizes.
Details on the obtained parameter distributions can be found in the
report mentioned below.
The test instances include also lengths of the berths in container
terminals present in the above ports.
Test Instances
The test instances were generated for a year timespan, with different
weekly intensities.
Since vessel arrivals are generated pseudo-randomly, the actual number
of arrivals may slightly differ from the product of the number of weeks
in a year and weekly intensity.
Instance File Format
Instance files are text files with data in the following order:
- 1st line: number of the ship calls at a port (n)
- lines 2 to n+1 - ship data (in the order): id (j),
ready time (r_j) in hours since the beginning of a year,
length (L_j), processing time (p_j) in hours,
weight (importance, e.g. cost of moving the containers w_j),
ship owner (o_j - can be ignored)
- line n+2: number of berths (m)
- line n+2 (the last line): berth lengths (\lambda_1,...,\lambda_m)
File Name Convention
portintensityninst#.txt
where:
port is an abbreviation of port name:
hb - Hamburg,
lh - Le Havre,
rt - Rotterdam,
sh - Shanghai,
si - Singapore;
intensity is the number of weekly ship arrivals;
n - just letter n;
inst# is the number of the instance.
The instances are collected in the following files
| # | instances | comments |
|---|---|---|
| 1. | intensity20.zip | intensity 20/week |
| 2. | intensity50.zip | intensity 50/week |
| 3. | intensity100.zip | intensity 100/week |
| 4. | intensity200.zip | intensity 200/week |
| 5. | intensity500.zip | intensity 500/week |
Technical Report
In this report parameters for container ship traffic model are provided.:
J.Wawrzyniak, M.Drozdowski, É.Sanlaville,
Ship Traffic Model for Container Terminals – Distributions and Parameters
Institute of Computing Science,
Poznan University of Technology,
Research Report RA-1/21, 2021.