Building large models for the generation of production-quality embedded software requires the development of a modeling style that guides and enforces model architecture, interface definition, modeling standards, and data management.
Models necessarily rely on external data and functionality to create an environment that allows initialization, trim, linearization, simulation, analysis, and code generation. This paper describes the fundamentals of how to define and manage parameters and signals within a model. It also discusses the implications of data management style on componentization, flexibility, readability, and code generation. Recommendations for models targeting embedded code generation for mission-critical and high-integrity systems are highlighted.
This paper was presented at the American Astronautical Society Conference.