Framework for Knowledge Aggregation with Subjective Logic in Multi-Agent Self-Adaptive Cyber-Physical Systems
This repository contains the source code and resources to perform Knowledge Aggregation with Subjective Logic in a ROS-based Simulated Self-Adaptive Multi-Robot System.
Academic Citations
Please cite the following paper when using this tool.
Petrovska, Ana, et al. "Knowledge aggregation with subjective logic in multi-agent self-adaptive cyber-physical systems." Proceedings of the IEEE/ACM 15th International Symposium on Software Engineering for Adaptive and Self-Managing Systems. 2020.
@inproceedings{DBLP:conf/icse/PetrovskaQGP20, author = {Ana Petrovska and Sergio Quijano and Ilias Gerostathopoulos and Alexander Pretschner}, editor = {Shinichi Honiden and Elisabetta Di Nitto and Radu Calinescu}, title = {Knowledge aggregation with subjective logic in multi-agent self-adaptive cyber-physical systems}, booktitle = {{SEAMS} '20: {IEEE/ACM} 15th International Symposium on Software Engineering for Adaptive and Self-Managing Systems, Seoul, Republic of Korea, 29 June - 3 July, 2020}, pages = {149--155}, publisher = {{ACM}}, year = {2020}, url = {https://doi.org/10.1145/3387939.3391600}, doi = {10.1145/3387939.3391600}, timestamp = {Tue, 29 Dec 2020 18:35:13 +0100}, biburl = {https://dblp.org/rec/conf/icse/PetrovskaQGP20.bib}, bibsource = {dblp computer science bibliography, https://dblp.org} }
Workspace Structure
The ros folder contains the ROS implementation of a Multi-Robot system. This ROS-based testbed consists of a set of Python and C++ ROS nodes that allow for an easy experimentation with the knowledge aggregation using the different available fusion operators in subjective logic.
The subjective_logic directory contains the source code that implements the formalisms of subjective logic; in particular, it provides an API to perform the aggregation of the knowledge of multiple agents using different fusion operators.
The folders themselves contain a more detailed explanation about their content. In particular, the ros folder incorporates a detailed step by step guide on how to install and run the system, also including architectural details of the system.
Authors
- Sergio Quijano, M.Sc. sergio.quijano.r@gmail.com
- Malte Neuss, M.Sc. malte.neuss@tum.de
- Ana Petrovska, M.Sc. ana.petrovska@tum.de
Acknowledgments
We want to express sincere gratitude to everyone who was somehow involved in the development of the tool: Guan Erijage, Shanin Yousfi, Kevin Hawryluk, David Bartuseck, Theo Beffart, Ricardo Cedillo, M. Ansab Shohab and Martin Büchner.