Development of artificial intelligence algorithms to estimate the state of health of lithium-ion batteries from cell to pack level

The electric vehicle market is growing strongly and phenomenal progress is underway, in particular in terms of extending the vehicle’s kilometric range . The traction battery is the central organ of these vehicles that dictates this range.
The lifespan of such batteries depends on several environmental factors such as temperature, battery state of charge and power demand. Diagnostic assessments require input on  the nature  of the lithium-ion battery technology, its different aging mechanisms and its configuration, especially the number of cells in series or parallel, i.e. factors that make the service lifetime extremely difficult to predict. To respond to the complexities surrounding this issue , an artificial intelligence approach, such as "Machine Learning", may well  be viable and is the subject of interest in this PhD project.
For this doctoral  project, the student will have access to existing test  databases (internal to IFPEN or open-source) gathering electrothermal measurements on several technologies of lithium-ion battery cells. Additional experiments using battery test facilities will also be carried out in a battery laboratory from single cells up to a module level. For data mining, artificial intelligence approaches including "machine learning" algorithms, will be employed.
Throughout this PhD, the doctoral student will be responsible for:
Carrying out an extensive literature survey, and keep on top of technological developments related to this field,
-    carrying out electrical measurements in the laboratory,
-    putting in place the means to  automate data-mining,
-    proposing new digital methods ,
-    writing scientific publications in addition to a PhD thesis.
This work requires an aptitude and willingness for processing and analyzing data through the use of  computer tools such as Python, R, Matlab, etc. Basic knowledge in electrochemistry and/or electrical engineering, some hands-on experimentation, as well as statistical analysis methods, would be requisite.

Keywords: Artificial intelligence, machine learning, batteries, aging, state of health

• Academic supervisor    Prof. VENET Pascal AMPERE UMR CNRS, Lyon 1 University. 
• Doctoral School    Ecole doctorale EEA de lyon, https://edeea.universite-lyon.fr/
• IFPEN supervisor     Dr MINGANT Rémy,ingénieur de recherche, Electrochemisty and Materials department,  Remy.mingant@ifpen.fr  
• PhD location    IFP Energies nouvelles, Solaize, France and AMPERE, Lyon 1 University, France and Siemens Digital Industries Software, Lyon, France  
• Duration and start date    3 years, starting in fourth quarter 2021
• Employer    Siemens Digital Industries Software, Lyon, France
• Academic requirements    Engineer or University Master degree in relevant disciplines 
• Language requirements    Fluency English, Basic knowledges in French