Pubblicazioni
| Autori: | Daniele Codetta Raiteri |
| Luigi Portinale | |
| Area Scientifica: | Artificial Intelligence |
| Diagnosis | |
| Uncertain Reasoning | |
| Probabilistic Graphical Models | |
| Dependability and Reliability | |
| Titolo: | ARPHA: an FDIR architecture for Autonomous Spacecrafts based on Dynamic Probabilistic Graphical Models |
| Apparso su: | TR-INF-2010-12-04-UNIPMN |
| Editore: | DiSIT, Computer Science Institute, UPO |
| Anno: | 2010 |
| Tipo Pubblicazione: | Technical Report |
| URL: | http://www.di.unipmn.it...R-INF-2010-12-04-UNIPMN.pdf |
| Sommario: | This paper introduces a formal architecture for on-board diagnosis, prognosis and recovery called ARPHA. ARPHA is designed as part of the ESA/ESTEC study called VERIFIM (Verification of Failure Impact by Model checking). The goal is to allow the design of an innovative on-board FDIR process for autonomous systems, able to deal with uncertain system/environment interactions, uncertain dynamic system evolution, partial observability and detection of recovery actions taking into account imminent failures. We show how the model needed by ARPHA can be built through a standard fault analysis phase, finally producing an extended version of a fault tree called EDFT; we discuss how EDFT can be adopted as a formal language to represent the needed FDIR knowledge, that can be compiled into a corresponding Dynamic Decision Network to be used for the analysis. We also discuss the software architecture we are implementing following this approach, where on-board FDIR can be implemented by exploiting on-line inference based on the junction tree approach typical of probabilistic graphical models. |