[en] The standard model (SM) of particle physics is a theory, describing three out of four fundamental forces. The prediction of the standard model are in good agreement with the results of measurement of the experiments at the CERN, the Fermi lab and other research establishments. Still, not all open questions of the particle physics could be answered with this model. For example, it is not possible to implement the fourth fundamental force, the gravity, into the SM. Another example are candidates for the dark matter, which couldn't described by the standard model, but cosmological experiments show that the contribution of the dark matter is about 25% of the universe. The supersymmetry, which introduce a symmetry between fermions and bosons, is regarded as one of the most promising solution for these open questions. In this theory, each new particle so-called supersymmetric particles is related to one particle in the standard model. If supersymmetry is realized in nature, one possible model of this symmetry is the mSUGRA model with R-parity conservation. In this model the lightest supersymmetric particle (LSP) is neutral and stable, so that it cannot be measured directly in the detector but indirectly over the missing transversal energy (etmiss). The search for new physics with the ATLAS experiments will start at the beginning of 2010 at the pp-accelerator LHC (providing pp-collision at √(s)=7-10 TeV at the startup phase) at a luminosity of L(t) =1 x 10³⁴ cm^-2s^-1. Because of the high data rate due to the 10⁸ readout channels of the ATLAS detector and a bunchcrossing rate of 40 MHz a trigger system is needed to reduce the date rate for storing. The choice of parameter for the trigger system is a compromise between the available bandwidth for the trigger rate and the selection efficiency for the physics events, because events with new physics are only in one of 10⁸ events expected. To fulfill the given requirements, a trigger system with three levels of date reduction is needed. The highest data reduction will take place one the first trigger level (Level-1 trigger). Within the scope of this work, an important contribution to the understanding of the property of the etmiss calculation is achieved for the Level-1 trigger. Furthermore new methods are presented to evaluate the etmiss-efficiency for standard model process and possible mSUGRA scenarios from recorded data. A maximum trigger rate of 100 Hz at a luminosity of L(t)=1 x 10³³ cm^-2s^-1 was specified for the optimization of the etmiss-trigger. For the trigger optimization, several simulation programs are used and for some of this programs dedicated software developing are needed. It is shown, that the discovery potential (for a signal significance of at least 5 sigma) is increased by combination of a etmiss-threshold with lepton and/or jet trigger threshold opposite the existing ATLAS-trigger menu on the Level-1 trigger up to 66 %. (orig.)