Implementation of a Model-Based Flight Controller on an ARINC 653 Operating System

Kurzfassung

The aerospace industry's increasing adoption of both Integrated Modular Avionics (IMA) and Model-Based Software Development (MBSD) creates a persistent integration challenge. While ARINC 653 provides a robust framework for partitioning software architectures, and MBSD facilitates verifiable application design, bridging the gap between these paradigms remains a complex, manual, and error-prone process. This is especially evident when managing InterPartition Communication (IPC) between heterogeneous partitions, implementing ARINC 653 application software in C-code or MATLAB Simulink. This paper presents the successful development and implementation of an extended toolchain and a flight controller software platform based on the SYSGO PikeOS Real-Time Operating System (RTOS) deployed on the DLR Small Avionics Computer (DSAC), which was developed by the German Aerospace Center (Deutsches Zentrum für Luft- und Raumfahrt, DLR). The work validates PikeOS as a feasible platform for implementing ARINC 653 IMA capabilities on the DSAC hardware, including the extensive use of external interfaces. The primary contribution is the significant expansion of SYSGOs Simulink integration toolchain that demonstrates the practical and efficient integration of MBSD into the PikeOS environment. A novel method for the programmatic, dynamic generation of Simulink adapter models was developed and applied. This key innovation automates the creation and management of IPC interface definitions, enabling seamless communication between the heterogeneous partition types. This sustainable approach reduces integration effort and mitigates critical errors caused by subsequent interface changes, providing a robust and functionally verified platform for the rapid deployment of complex, model-based avionics applications. Through the full automation of previously manual steps, the toolchain is industrialized and made suitable for modern Continuous Integration and Continuous Deployment (CI/CD) workflows.