Artificial Intelligence

Computer Science | Faculty of Engineering, LTH

Denna sida på svenska This page in English

Finished examination projects

Reports accessible below are written in English, unless otherwise noted.








  • Johan Hovold:
    On a Semantics for Active Logic. (short summary /in Swedish/, abstract, BiBTeX entry.)
  • Johan Wessfeldt:
    AI-Middleware: Datatstrukturer för vägsökning med A*-metoden för en klass av strategispel.
  • Henning Ask:
    AI-Middleware: Undersöningar av användning av smart terräng i strategispel.
  • Tim Portnoff:
    Reinforcement Learning with Continuous States, Based on Context-Q.
  • Björn Nilsson:
    Konstruktion av dynamiska korsord.


  • Stefan Bie and Johan Persson:
    Behaviour-Based Control of the ERS-7 AIBO Robot (full report). This work has been done within the Team Sweden project.
  • Iñaki Navarro Oiza:
    Reactive Cooperation of AIBO robots (summary and full report). This work has been done within the Team Sweden project.
  • Jonas Isberg:
    Using Interactive Computer Games for AI Research. Case Study on WarCraft III: summary (PDF), full report (PDF) and WarCraft III program pieces (.tar.gz).
  • Fredrik Winge:
    Sonar för en Khepera: summary (PDF), full report (PDF) and a piece of code.
  • Sonia Fabre:
    Automatic theorem proving in Labelled Deduction Systems (presentation slides (PDF) and, CommonLisp code).
  • Erik Grönlund:
    A Tool for Case-Based Reasoning.
  • David Clauss and Johan Sjöström:
    Modelling of water flows to power basins by recurrent neural nets.




  • Håkan Johansson:
    Robotic courses based on LEGO (a PDF file).
  • Henrik Öhman:
    Från myror till samarbetande robotar, (report, written in Swedish, in PDF or PostScript, plus the C code of the program described there).
  • Niklas Persson and Christian Rahm:
    Niklas and Christian have looked at creating teams of RoboCup agents. Niklas has looked at the decision trees as a medium for implementing proper decision-making (his report, written in Swedish, in HTML, and his code), while Christian has tried to implement neural learning of basic skills, with rather discouraging results (His report, written in Swedish, in HTML, and his teams: with hardcoded skills and with learning ones.)