par Massink, Mieke;Brambilla, Manuele 
;Latella, Diego;Dorigo, Marco ;Birattari, Mauro
Référence Swarm Intelligence, 7, 2-3, page (201-228)
Publication Publié, 2013
;Latella, Diego;Dorigo, Marco ;Birattari, Mauro Référence Swarm Intelligence, 7, 2-3, page (201-228)
Publication Publié, 2013
Article révisé par les pairs
| Résumé : | In this paper we analyse a swarm robotics system using Bio-PEPA. Bio-PEPA is a process algebra language originally developed to analyse biochemical systems. A swarm robotics system can be analysed at two levels: the macroscopic level, to study the collective behaviour of the system, and the microscopic level, to study the robot-to-robot and robot-to-environment interactions. In general, multiple models are necessary to analyse a system at different levels. However, developing multiple models increases the effort needed to analyse a system and raises issues about the consistency of the results. Bio-PEPA, instead, allows the researcher to perform stochastic simulation, fluid flow (ODE) analysis and statistical model checking using a single description, reducing the effort necessary to perform the analysis and ensuring consistency between the results. Bio-PEPA is well suited for swarm robotics systems: by using Bio-PEPA it is possible to model distributed systems and their space-time characteristics in a natural way. We validate our approach by modelling a collective decision-making behaviour. © 2013 Springer Science+Business Media New York. |
