Section: New Results

Predictive maintenance for an optronic equipment

Participants : Camille Baysse, Benoîte de Saporta, Anne Gégout-Petit, Jérôme Sarraco.

After the problem of detection of a degraded state, we have tackled the problem of predictive maintenance for an optronic equipment. For this we model the state of the system by a PDMP (state with three possible values and cumulative time of use). In this framework, we reformulate the problem of maintenance of optimization in an optimal stopping problem maximizing a criteria about time of use without failure. In this framework, we can use known results developed in the CQFD team on optimal control [8] , [23] . We have extensively studied the problem with simulated data, computed grid of quantization and optimal policy for the real problem. This results will be implemented by Thales in HUMS of optronic equipment.

This work was presented in an national conference [38] and an abstract is accepted for publication in an international conference with papers.