Lockheed Martin has implemented a composable design methodology in the update of its A2100 Satellite product line. The composable design methodology leverages model-based systems engineering language and tools to formalize allowable architectural choices within the product line, limiting the decision points and decision options available to system architects, as well as imposing constraints on how those decisions may be combined. Even in the context of those limitations, a complex system can have a large design space, required to support flexibility for widely varying missions, payloads, and customers. This can result in many valid architecture configurations that satisfy a given set of mission requirements. When faced with identifying the preferred satellite architecture for a given mission, a system architect may often resort to the familiar rather than evaluating all potential configurations for the optimal balance of cost, schedule, and performance. By leveraging optimization techniques with a Composable System Reference Architecture, Lockheed Martin enables a system architect to rapidly identify a subset of configurations that form an optimal pareto front based on performance and cost objectives. The system architect can choose a configuration that best achieves all identified objectives, while retaining flexibility to prioritize among competing objectives in a manner most appropriate for their program. This paper shows that the composable model suite developed by Lockheed Martin supports repeated analysis of optimal architecture configurations to satisfy varying mission and customer applications from program to program.
Using Optimization to Exploit a Composable Satellite Product Line Architecture
MBE: Model Based Engineering, MBSE: Model Based Systems Engineering, MDAO: Multi-Disciplinary Analysis and Optimization
Abstract