A methodology for system-of-systems design in support of the engineering team

摘要

Space missions have experienced a trend of increasing complexity in the last decades, resulting in the design of very complex systems formed by many elements and sub-elements working together to meet the requirements. In a classical approach, especially in a company environment, the two steps of design-space exploration and optimization are usually performed by experts inferring on major phenomena, making assumptions and doing some trial-and-error runs on the available mathematical models. This is done especially in the very early design phases where most of the costs are locked-in. With the objective of supporting the engineering team and the decision-makers during the design of complex systems, the authors developed a modelling framework for a particular category of complex, coupled space systems called System-of-Systems. Once modelled, the System-of-Systems is solved using a computationally cheap parametric methodology, named the mixed-hypercube approach, based on the utilization of a particular type of fractional factorial design-of-experiments, and analysis of the results via global sensitivity analysis and response surfaces. As an applicative example, a system-of-systems of a hypothetical human space exploration scenario for the support of a manned lunar base is presented. The results demonstrate that using the mixed-hypercube to sample the design space, an optimal solution is reached with a limited computational effort, providing support to the engineering team and decision makers thanks to sensitivity and robustness information. The analysis of the system-of-systems model that was implemented shows that the logistic support of a human outpost on the Moon for 15 years is still feasible with currently available launcher classes. The results presented in this paper have been obtained in cooperation with Thales Alenia Space—Italy, in the framework of a regional programme called STEPS.