Development of a multiobjective methodology for layout optimization of equipments in artificial satellites

Ana Paula Curty Cuco, Fabiano Luis de Sousa, Antônio José da Silva Neto (Instituto Nacional de Pesquisas Espaciais, Universidade do Estado do Rio de Janeiro)

CHALLENGE - The National Institute for Space Research (INPE) needed to use a multi-objective and multidisciplinary design to create an artificial satellite that would have a structure supporting the many pieces of equipment placed inside. Placement was required to consider the position to the center of the mass of the satellite, heat dissipation of electronic equipment, the need to group them by subsystem, and the alignment requirements and moments of inertia.  











INPE aimed for the following objectives and constraints:

  • minimize distance between the center of mass of the layout found numerically and the center of mass of the satellite
  • minimize the thermal power density by the electronics on the panels
  • minimize the angles between the principal axes of inertia of the layout obtained numerically and the reference axes.
  • all components must be contained within the module
  • no overlap or interference between equipment and the satellite module or the equipment itself

SOLUTION - Researchers formulated a multi-objective optimization problem using the Pareto optimality criterion that would obtain the best design while considering the design's constraints. Even while the constraints left starting with an initial unfeasible count of designs, it soon became possible to find feasible solutions after generations. Additionally,learned that further investigation was required on the positioning of similar equipment.


BENEFITS - The work flow incorporated Excel, Matlab and SolidWorks into modeFRONTIER and concluded that trade-off solutions were possible and better than the original manual design. After 200 generations of optimized designs, despite the imposing constraints, the results provided by modeFRONTIER helped move past infeasible designs and find a solution. Additionally, that the proposed methodology is very promising for the conceptual design phase and the case study is valid as a conceptual feasibility test.