ALFRED2: An Abductive Logic FRamEwork for Distributed Diagnosis

Last modified by Andrea Omicini on 16/06/2021 22:15

Marco Battarra

The aim of the project is to build a multi-agent framework for general purpose diagnosis, based on a soft computing abductive approach. Such framework should be more practical and more effective in diagnosis applications than the existing ones. It must be able to guide a user through the process of diagnosis, from the initial observations to a final diagnosis. The final diagnosis must be checked and corroborated by means of appropriate tests. To reach this aim, the objectives are:
  • To research in scientific literature the main abductive approaches and to study their possible applications in fault diagnosis.
  • To detect the practical limitations of the existing approaches and to devise a soft computing approach to increase effectiveness.
  • To develop a multi-agent framework based on the approach pre- viously devised.
  • To compare the behaviour of the developed framework with the behaviour of the frameworks identified in the research stage. Such evaluation must be based on the same case study of fault diagnosis.

Thesis Data

2008 © aliCE Research Group @ DEIS, Alma Mater Studiorum-Università di Bologna