Abstract: |
In this paper, a new hybrid evolutionary strategy is used to the inverse design of a spacecraft thermal control system. This hybrid strategy comprises one step of global, stochastic search, using the Generalized Extremal Optimization (GEO) algorithm, followed by a local, deterministic refinement minimization step with the EXTREM optimization routine. The GEO algorithm is a novel global search meta-heuristic, based on a model of natural evolution, while EXTREM uses Powell’s method of conjugate directions. This approach is applied to the inverse design of a spacecraft thermal control system, considering different critical, on-orbit operational conditions. Numerical results show that the hybrid approach improves the design solution in terms of the value of the objective function, while GEO yields non-intuitive efficient solutions in the sense they were unlikely to come out through the classical, “manual” design procedure. |