DC plasma spraying is due to a non-linear problem with several variables. Various calculation models have been developed to predict the jet characteristics, the "in-flight" particle behavior, or the particle shaping mechanism. On the one hand, these models are very complex due to their explicit nature, but on the other hand, they are very special because they relate to very specific facts of the overall process. This article offers a global approach based on an implicit description and incorporating artificial neural networks (ANN). ANN is an implicit model that not only takes into account the complexity of the problem but also the interaction of the individual parameters and measurement errors. The global concept and the protocols to be considered are presented and explained in relation to the direct current plasma spray process. The article describes an example having the purpose to link processing parameters with in-flight particles characteristics". Paper includes a German-language abstract.