Obtaining a relevant model for a complex system is striven for in the Health Management community, as it allows to precisely monitor the health state of the system. This article introduces a method to obtain a hybrid model under the Heterogeneous Petri Net formalism for a system, using only data. The method is comprised of two steps, the learning of the Discrete Event System (DES) structure of the system using a clustering algorithm, then the learning of the continuous dynamics contained in the system and its various functioning modes using two regression algorithms. The method is applied on an academic example.