Facilitating Topological Design Decisions in Early Mission Design Phases With Automated Knowledge Graph Generation by Means of Graph-Based Design Languages

Satellite design projects in the new space age are facing the challenge of increased complexity and increased number of deliverables while the development time must be reduced in order to remain competitive. To handle the complexity of modern satellite systems, model-based systems engineering (MBSE) introduces visual models to further the understanding of complex system relationships, aiding the engineering team in the design and integration of components and subsystems. Graph-based design languages enable an automatic execution of MBSE models. A central design graph, a type of knowledge graph, is the central source of truth along the entire design process, ensuring consistency of the design over all development phases. One advantage of this executable graph-based data model is the ability to automatically create consistent but topologically different designs. These topological variants include different configurations of the satellites subsystems, such as the number of panels, their mounting on the satellite, available sensors or number of actuators. For an optimal design of each subsystem, the number and configuration of components depends on the mission requirements and the operation modes. Each operation mode fulfills a function, defining a required set of components, which in turn defines the power demand of the mode, and therefore the power requirements of the power system. A change in an operation mode, such as its duration or its required power, necessitates a redesign of each subsystem. Conversely, a design constraint on a subsystem requires a reevaluation of the operation modes. This work explores the interaction between subsystem design and operation modes in an executable process, automatically deriving operation modes from mission design parameters and generating the design of the satellite including different configurations of solar arrays. These configurations are generated based on different mission parameters such as orbit height and payload resolution. This approach enables engineering teams to reduce design time while exploring an increased number of different design topologies.