Thermal parameters identification in the correlation of spacecraft thermal models against thermal test results

Temperatures predicted by the Thermal Mathematical Models (TMMs) used in the thermal control design of spacecraft, usually present differences with the values measured during the thermal test campaign. Therefore, the TMMs must be correlated with the thermal tests to reduce these differences to admissible values. This task can be addressed in an automatized way considering the correlation as an optimization problem, where the differences between predicted and measured temperatures are minimized. This is achieved modifying the values assigned to some parameters used in the TMMs. The main drawback of this approximation is the risk of loosing the physical sense of some model parameters. The reason is that the thermal inverse problem, that is, calculate the thermal parameters that produce a specific temperature distribution, often has not a unique solution. A methodology of automatized correlation to calculate the correct values of the model parameters, in the sense that they maintain its physical interpretation, is presented in this article. The key point relies in setting up an overdetermined system of equations. The expression to calculate the minimum number of load cases required, is developed, and several case studies are presented to validate the proposed methodology. A gradient based public available set of subroutines (TOLMIN) has been used as optimization algorithm.