Trainee Project
Title:
Evaluations of machine learning techniques based on convolution neural networks and reinforcement learning techniques for PHM and predictive maintenance
Dates:
2019/06/03 - 2019/08/30
Supervisor(s): 
Description:
Evaluations of machine learning techniques based on convolution neural networks and reinforcement learning techniques for PHM and predictive
maintenance



Step 1: Scientific bibliography that will present the techniques in a general way and the different versions or evolutions if there are any. A focus will
be placed on the application of these techniques for PHM and more particularly for early detection, diagnosis and prognosis.

Step 2: State of the art commercial solutions in the field of PHM and also the tools / libraries paid or royalty-free to implement these techniques. The
different solutions will be compared according to criteria to be defined

Step 3: Putting a test dataset into practice and evaluating performance against current PREDICT techniques

Step 4: Assessment, Synthesis and Report
Department(s): 
Eco-Technic systems engineering