The increasing pressure drop of filter media during the capture of solid particles is directly correlated to the energy consumption. It is largely based on the microstructure of the media, especially with regard to the selected porosity gradient. The development of new filter media optimized for separation efficiency and energy consumption is based mainly on experimental work. The use of evolutionary algorithms inspired by nature is of particular interest for the optimization of filter media, as a large number of non-linear influencing factors can be processed. However, a large number of simulations is required for successful application in order to reproduce the natural process of mutation and individual selection. Most numerical simulations of filtration properties are limited due to the high computing power required, making them unsuitable in this context. Therefore, a 1D simulation with low computing power requirements was used in this paper to simulate the loading of a filter medium with particles. This approach enabled the rapid calculation of the loading of different filter media structures, and was coupled with an evolutionary algorithm with the aim of optimizing the porosity gradient. The algorithm was tested on different particle size distributions, which represented different operating conditions. The porosity gradients of the optimized filters was in the form of an inverse square function, leading to a much more homogenous, microscopic loading with particles. Due to the significantly improved utilization of the filter depth compared to a reference medium, a much slower increase in pressure difference could be obtained with a similar filtration efficiency, resulting in increased quality factors of 30% and higher. The optimization of an operating point took about 24 hours on a standard desktop PC and could be extended to other structural parameters of the filter medium apart from the porosity gradient. This shows the potential for a broad application in the selection and optimization of filter media.