This diploma thesis deals with a special Human Resource Allocation Problem, in which a given amount of resources has to be distributed among an also given amount of stations. For each resource it is known, which performance it can achieve on each of the stations. The stations can also be connected in such a way, that a station can only achieve its full performance, if it gets enough input from its predecessors and if it is also able to pass on its output to its successors. The stations and their connections form a network and the maximal flow through this network determines the performance of an allocation. The model is constructed in such a way, that the capacities of the edges depend on the allocation of the resources to the stations. Buffers between stations, preallocation of resources and minimum performance constraints for stations can also be considered by this model. Finding an allocation which produces the best maximum flow is the main goal of this thesis. To find the best – or at least good – allocations, I involved genetic algorithms. I applied this approach to several problems, thereby using variants of genetic algorithms with different parameter sets as well as selection- and crossover strategies.