Comparison of Multidisciplinary Design Optimization Architectures for the design of Distributed Space Systems

Advancement in satellite technology, and the ability to mass-produce cost-effective small satellites has created a compelling interest in Distributed Space System (DSS), such as Low Earth Orbit (LEO) satellite constellations. Optimization of DSS is a complex Multidisciplinary Design Optimization (MDO) problem involving a large number of variables and coupling relations. This paper focuses on comparing three different MDO architectures for a DSS design problem. Initially, an overview of the constellation model, the subsystems model, and the coupling relationships between the subsystems and the constellation are provided. The modelling of the subsystems and the constellation configuration are carried out in OpenMDAO. Later, three monolithic MDO architectures, namely, Individual Discipline Feasible (IDF), Simultaneous Analysis and Design (SAND) and Multidisciplinary Feasible (MDF) are compared by implementing them to the developed DSS model. The results indicate IDF outperforms the rest of the architectures for the conceptual design of DSS. The optimum objective function obtained by IDF is 1% lower than SAND and 7% lower than MDF. While the functional evaluation required for IDF is 50% lower than SAND and 90% lower than MDF.