direkt zum Inhalt springen

direkt zum Hauptnavigationsmenü

Sie sind hier

TU Berlin

Page Content

MCAS – Model Check Methodology for Avionic Software

STEMME S15 Flugversuchsträger aus den Projekten LAPAZ und LAPAZ II
Lupe

MCAS is a joint research project to design a verification process for the model-based development of safety-critical avionics software. The project is supported by the Federal Ministry for Economic Affairs and Energy (BMWi) as part of the Central Innovation Program for Small and Medium-sized Enterprises (ZIM). Its duration is 33 months (01.08.2015 - 31.04.2018)

Introduction

Electronic control systems for vehicles contain functions that are both complex and safety-critical. Examples are electronic flight control systems for aircraft or advanced driver assistance systems in the automotive industry. The certification of such systems requires a well-structured development process that is governed by strict, industry-specific regulations. The risk, effort and ultimately cost of the development process can be reduced significantly, by employing a model-based software development process.
Model-based software development is characterized by the implementation of an executable model of the software that is being developed. There are two advantages:

  1. The software model can be used to generate code automatically,
  2. The developed functions can be validated at an early stage with simulation.

The guideline RTCA DO‐331 "Model‐Based Development and Verification Supplement to DO‐178C and DO‐278“ released in December 2011 opens up the possibility to employ model-based software development in the development of avionic software – particularly in the certification process – and requires verification of the utilized models. There is a lack of strict modeling rules and efficient methods to verify the compliance of models to that rules for common development tools. The required effort to develop guidelines and verification methods particularly hinders small and medium sized enterprises in developing avionics software for CS-23 aircraft.

Goals

The MCAS project aims at reducing the required effort for development and certification of avionic software, to facilitate the development and implementation of electronic flight control systems for aircraft in the CS-23 segment. The objective is the development of a verification process that ensures the compliance of the model-based development process of avionic software to the requirements set forth in RTCA DO-331. Possible errors should be identified and rectified during the modelling phase by frequent, automated, static analysis of software models throughout the course of development. The practical implementation of the verification process is based on MATLAB®/ Simulink®/ Stateflow®/ TargetLink® as a commonly used tool chain for model-based development of embedded software systems.

Project plan

AP100
Classification of models, establishment of requirements
Deriving specific requirements for specification models and design models; Analysis of demand for higher quality checks for verifying models.
AP200
Steps for verifying the required properties
Design of model patterns for specification and design models and prototypical check algorithms for Simulink®/Stateflow®.
AP300
Design of integrated check methodology
Integration of all test steps relevant for certification into a three-layered verification method encompassing checks of specification models, design models and the generated source code.
AP400
Validation and evaluation of the verification method
Evaluation and validation of the integrated check methodology using existing complex avionic software models especially of flight control systems.

Tasks of TU Berlin

The Department for Flight Mechanics, Flight Control and Aeroelasticity is responsible for work package AP300 and parts of work packages AP100 and AP400. The tasks are:

  1. Development of a catalogue of requirements for software models according to RTCA DO-331 (AP100).
  2. Definition and implementation of methods that enables verification of specification models, design models and generated source code (AP300).
  3. Evaluation and validation of the test method using existing complex models (AP400).

Project Partners

  • Model Engineering Solutions GmbH
  • Department for Flight Mechanics, Flight Control and Aeroelasticity (FMRA) at the Technical University of Berlin

Funding

Lupe

Zusatzinformationen / Extras

Quick Access:

Schnellnavigation zur Seite über Nummerneingabe

Auxiliary Functions

Dipl.-Ing. Georg Walde
+49 (0)30 314-21333
F-Building
Room 341