This paper intends to present a modeling method that combines neural networks and Petri nets, a method we will term as NPN (Neural Petri Net). The proposed method the advantages of a manner of modeling based both on Pnets and on neural networks. NPN permits a much better modeling of some complex processes we shall describe as possible. The utilization of this NPN method permits the inclusion of neural networks in Petri which provides solutions for the modeling of systems with artificial intelligence or for systems that have a complex evolution.