Application study of a correction method for a spacecraft thermal model with a Monte-Carlo hybrid algorithm

Abstract

The correction of a thermal model for a thermally controlled satellite in ground test conditions is studied using a Monte Carlo hybrid algorithm. First, the global and local parameters are summarized according to sensitivity analyses on uncertain parameters, and then the model correction is treated as a parameter optimization problem to be solved with a hybrid algorithm. Finally, the correction of the thermal model is completed using a layered correction method. The sensitivity analysis showed that the effective emissivities across the multi-layer insulation (MLI) and the emissivities of the thermal control coating are global parameters, while the contact heat transfer coefficients are local parameters. After correction, the deviations between the calculated and test values were all within ±3°C. The final results prove that the method in this study is superior to traditional methods and satisfies the requirements for thermal model correction.

On Using Quasi-Newton Algorithms of the Broyden Class for Model-to-Test Correlation

“…In contrast to most correlation methods developed (Jouffroy, 2007;De Palo et al, 2011;Momayez et al, 2009;Harvatine and DeMauro, 1994;Roscher, 2006;van Zijl, 2013;WenLong et al, 2011;Mareschi et al, 2005), this method will not attempt to minimize the length of the vector ||F(p) -t mes ||, but searches a root of the equation system:…”

“…Many methods have been developed and analyzed to perform model-to-measurement correlation (Jouffroy, 2007;De Palo et al, 2011;Momayez et al, 2009;Harvatine and De Mauro, 1994;Roscher, 2006;van Zijl, 2013;WenLong et al, 2011;Mareschi et al, 2005). Most methods are based on stochastic optimization algorithms and often require several hundred iterations to converge.…”

Performance of Optimization Algorithms in the Model Fitting of the Multi-Scale Numerical Simulation of Ductile Iron Solidification

“…For example Dudon [11] use the Latin hypercube sampling (LHS), the self-adaptive evolution (SAE) and the branch and bound methods; Beck [12] use the adaptive particle swarm optimization (APSO); Van Zijl [13] use the Monte Carlo, the genetic algorithm (GA) and the APSO methods; Trinoga and Frey et al [14,15] use the simulated annealing (SA), the threshold accepting (TA) and the GA; Anglada et al work mainly with GA [16,17] but also made some comparisons with Klement algorithms [18] and with the deterministic algorithms TOLMIN, NEWUOA, BOBYQA and LINCOA [19]. In addition, works about the use of hybrid algorithms combining deterministic and stochastic methods can also be found, as the works of De Palo [20] where the downhill simplex is combined with the LHS or the work of Cheng et al [21] where the Broyden-Fletcher-Goldfarb-Shanno (BFGS) is combined with the Monte Carlo method.…”