Numerical optimization of ESA’s messenger space mission benchmark

Abstract

The design and optimization of interplanetary space mission trajectories is known to be a difficult challenge. The trajectory of the Messenger mission (launched by NASA in 2004) is one of the most complex ones ever created. The European Space Agency (ESA) makes available a numerical optimization benchmark which resembles an accurate model of Messengers full mission trajectory. This contribution presents an optimization approach which is capable to (robustly) solve ESA’s Messenger full mission benchmark to its putative global solution within 24 h run time on a moderate sized computer cluster. The considered algorithm, named MXHPC, is a parallelization framework for the MIDACO optimization algorithm which is an evolutionary method particularly suited for space trajectory design. The presented results demonstrate the effectiveness of evolutionary computing for complex real-world problems which have been previously considered intractable.