Datenbestand vom 29. November 2024
Verlag Dr. Hut GmbH Sternstr. 18 80538 München Tel: 0175 / 9263392 Mo - Fr, 9 - 12 Uhr
aktualisiert am 29. November 2024
978-3-8439-4608-7, Reihe Informationstechnik
Marc Riedlinger Architecture of an Adaptive Avionics Platform Supervisor
177 Seiten, Dissertation Universität Stuttgart (2019), Softcover, B5
The complexity of today's avionics is caused by its extensive functionality, safety criticality, real-time behaviour and multiple operating modes in order to cover all life cycle-functions. Most architectures are distributed, highly redundant and integrated. This complexity, in combination with strict certification procedures, leads to an extensive development process and generates unchangeable designs. In some aircraft system domains, the resulting inflexible life cycle of avionics contradicts the commercial requirement for changeability. E.g. for cabin management systems (CMSs), changeability includes cabin customization before initial aircraft delivery, in-service cabin layout changes as well as functional upgrades on retrofits.
State-of-the-art CMS realize changeability with configurable system architectures. For configuration, parameters with a low abstraction level are manually set. This approach provides a vast change domain, but the change process itself is very time consuming and error-prone. To overcome this issue, this thesis deals with an adaptive avionics approach. It allows the creation of Adaptive Systems of Integrated Systems (ASIS). An ASIS is able to self-configure with minimal human interaction to changes during an on-ground organization phase. Changes may be avionics scaling (changing the avionics structure including safety) and / or avionics modification (changing the functionality).
This thesis introduces an architecture for the platform supervisor service, which is a significant part of the failure-, redundancy- and operating management of an ASIS. Failure management is the detection and isolation of faulty resources, e.g. computing modules, network units or peripherals. Redundancy management is the abstraction of redundancy towards applications and the optimized utilization of resources. Operating management is the coordination and enforcement of mode-dependant behaviour of ASIS components, such as test- or maintenance-modes.