Keynote speakers

Wednesday, August 28th

« Is prognosis of failure of Electrical Machines and Drives feasible and useful ? », Elias Strangas, Professor, Michigan State University

 

E. Strangas heads the Electrical Machines and Drives Laboratory at Michigan State University. He has been working on the design, control, and fault diagnosis of electrical machines and drives.  His students and he, for the last twenty years, have been focusing on faults in the power electronics, electrical machines and related mechanical systems and the prognosis of their failures. He was the General Chair of the 2017 edition of SDEMPED.

 

Whether explicitly or implicitly, Fault Diagnosis is directly connected to Failure Prognosis, necessary to decide on a course of action:  The options range from inaction, to mitigation, to emergency shutdown. Prognosis has to be timely and accurate to minimize the probability of a catastrophe and costs, and for that, it requires a detailed understanding and modeling of the deterioration process. This is not always done perfectly, so we are left with a process that we have to understand its limitations and uses. There are cases where prognosis can be implemented accurately and on time, while others when the time available to act is comparable or shorter than the lower estimate of the Remaining Useful Life.

We will touch on some of these cases, and identify what should be the next steps in the path to developing reliable and useful failure prognosis.

 

Thursday, August 29th

“Evolution of space launchers health management systems: from current status to future developments“,
Philippe Saunois, Serge Le Gonidec, Ariane group

Aim:To show and illustrate the HMS problematic for Space Launcher propulsive System

 

In the frame of the space launcher evolution, particularly due to the reusable reflections, the Health Monitoring/Management System is seen as fundamental topic. It is proposed a certain photo of our current status and the potential evolution of this technology. After description of our technical problematic, the overview will allow to highlight the technical need, the constraints and some solutions already evaluated. An illustration of the potential future will be shown …

 

A short summary:

  • Context :

-          What about reusable engine ?

-          Health Monitoring for what ?

-          Failure mode engine generic list

  • A little historic based on major activities:
  • HMS : A multidisciplinary challenge

-          Major thematics

-          Some Requirements

  • HMS: an automatized process of treatment

-          some strategies for some functional needs

  • Platform and demonstrations (some typical results)
  • The Health Control Interaction and Automatic fault tolerant control
  • Conclusion & challenges

 

Serge Le Gonidec was born in France, in 1964. He started its carrier in 1985 at the  “Société Européenne de Propulsion” in a support team “control, measurement and command system” on the engine test benches of the Ariane launcher.

 

Early in 1990, the Ariane 5 program gave him the opportunity to work on the test bench control engineering (flow control, tank pressurization, turbine speed, etc.). These new A5 test benches allowed introducing the digital control and its generalizationfor the engine or component tests under a high required safety.

 

In 1994, he rejoins the team of System department. Its activities are oriented modeling development and stability analysis for system developments or for the support for understanding the anomalies.

 

After 1997, he contributed to the engine design (e.g. Vinci) or evolution (e.g. Vulcain2) in introducing the Multi-variable Predicted Control. In the same time, in relation with ESA or CNES, he worked on R&T activities dedicated to the engine control and the Health Monitoring System.

 

These activities was a support to the future launchers preparation and have contributed to opening the way to the reusable launchers, control and monitoring interaction / IVHM / AFTC, …

 

In the frame of the company « ArianeGROUP » entity, he is currently member of the General Inspection team.

 

Philippe Saunois started its carrier in 1996 at Aerospatiale in a team dealing with Ariane 5 thrust vector control engineering. Since then, he has been working on main European launcher development programs in the field of mechatronics, multidisciplinary modelling and controlled systems. In the frame of the company « ArianeGROUP », his activities mainly cover launcher thrust vector control and attitude control applications, controlled ground means and R&T projects, with a special interest on study cases coupling mechanical dynamics, control and mechatronics.

 

Friday, August 30th

From thermal to electric powertrain: the next automotive diagnosis challenge",
M. THOMAS, Artificial Intelligence Team Leader, Innovation and Technology Department, ACTIA (Automotive company)

Jérôme Thomas studied computer science and artificial intelligence at Paris VI University. He obtained his master's degree in Artificial Intelligence and Pattern Recognition in 1991. In 1996 he defended his doctoral dissertation on machine learning and knowledge acquisition. He joined the ACTIA group in 2000 as a research engineer. His work focused on diagnostic methods applied to automotive. Since 2005, he is the head of the research and Artificial Intelligence team for ACTIA Automotive, the scope of this team extends from diagnosis/prognosis using data or model driven approaches to embedded computer vision using deep learning approaches.

Contact

SDEMPED secretariat

6 allée Emile Monso - BP 34038

31029 Toulouse Cedex 4

sdemped2019(at)inp-toulouse.fr

05 34 32 31 19