Understanding the Relationship between Lock Complex Effectiveness and System Performance: A Study of the Volkerak Complex

Welcome to the GitHub page where I share my code for the input & output analysis of my graduation project! Here is the digitial version of my report, if you would like to take a closer look into my motivation for this analysis or the discussion I motivated from this work.

Here is a short description of content in each notebook:

data_cleaning.ipynb (Data Cleaning & Initial Exploration)

• First look into the operational logs and ship classes
  • unallowed passages,
  • periods of renovation in lock chambers,
  • trends in ship characteristics,
  • passage times and lockages
• Cleaning the data to get it ready for the prediction.
• Inspecting malfunctions in the complex.

validation_locking_regimes.ipynb

Visualizing number of ships per lockage for different locking regimes, comparing them with the real observation.

arrival_analysis_prediction.ipynb (Analysis and Prediction of Arrivals)

• Identifying arrival pattern
• Work on seasonality
  • Observing seasonality over the years and per ship class
  • Clustering ship classes based on seasonality
  • Fitting SARMA/ARMA models to predict ship arrivals per ship class per direction
• Selecting the week to focus as baseline fleet mix
• Modifying the baseline fleet mix based on IMA, to create the IMA-driven fleet mix

stress_testing.ipynb

Observing waiting times and emissions under changing fleet intensity and lock condition scenarios.

univariate_<factor_explored>.ipynb

There are four factors on which univariate experiments are conducted:

1. Chamber priority
2. Inspection duration
3. Inspection start
4. Traffic range These notebooks generate boxplots of average waiting times and number of levelings for the univariate experiments.

---

I would appreciate your feedback on how I can improve this repository!