Le 10/02/2021 par jpuchinger :

Coordinated Shared Autonomous Electric Vehicles and Power Grid charging scheduling under uncertainty: a robust optimization approach
Supervisors: Jakob Puchinger, Adam Abdin
Laboratoire Génie Industriel / Future Cities Lab. (Ecole Centrale Pékin, CentraleSupélec)

Profile of candidate: M2 level master student with a specialization in Operations Research / Computer Science / Industrial Engineering / Applied Mathematics or a closely related field.

Location of internship: Laboratory of Industrial Engineering, CentraleSupelec, Université Paris-Saclay. 3 rue Joliot Curie, 92290, Gif-sur-Yvette, France.
Period: 6 months starting from the beginning of April 2021
Application: Candidates are kindly invited to send us their applications to jakob.puchinger@centralesupelec.fr and adam.abdin@centralesupelec.fr

including:
- An up to date CV indicating their experiences and competences.
- A motivation letter.
- A separate document with relevant courses / experiences / projects related to the topic proposed or the method indicated.

Description:
The aim of this research project is to leverage an emerging framework for handling problems of decision making under uncertainty, called distributionally robust optimization (DRO), to address the problem of coordinated shared autonomous electric vehicles (SAEV) and power grid charging scheduling.

The objective of the internship is to work on this problem starting from an existing literature review on the topic of coordinated SAEV and power grid charging schedule, developing the relevant model within the DRO framework and use it to arrive at useful insights to stakeholders and policymakers regarding the optimal coordination between the SAEV and power grid in terms of charging schedule. In particular, the candidate is expected to:
1. Complete the relevant state-of-the-art on the coordinated SAEV and power grid optimal charging scheduling, if needed.
2. Conceptualize the proper uncertainty / ambiguity sets to handle the relevant uncertain parameters within the DRO framework.
3. Integrate the uncertainty / ambiguity sets into the optimization problem and derive a solvable and scalable tractable form.
4. If the problem is computationally challenging to scale for large sized instances, work on developing a proper heuristic approach to solve the problem and validate the quality of its solution.