The proposed thesis work is part of the development of a technique to determine the actual and future state of a helicopter. This is a key issue in defining dynamic maintenance operations adapted to the actual condition and operational use of helicopters. Indeed, today it should be highlighted that the cost of the maintenance of a helicopter represents approximately one third of the costs that customers receive during the operation of their helicopter. Several approaches aimed at improving maintenance and reducing these costs have resulted in the development of algorithms, methods, tools,... to anticipate failure, malfunction or detect failure on a complex system.