‹ Back to the directory

Sophie Demassey

Sophie Demassey

Lecturer

Center · CMA

Discipline(s)
Applied Mathematics, Computer Science, Energy, Thermal Sciences, Signal, Image, Automatic Control, Robotics and Industrial Engineering
Topic(s)
Adaptation, AI Impact, CO2, Computer Science, Energy Efficiency, Grid/Network, Innovation and Design, Logistics, Maintenance and Durability, Managing Uncertainty, Nuclear, Public Policy, Renewable Energy, Scenario-Modelling, Sustainable Development, Water

Biography

Sophie Demassey holds MSc degrees in fundamental mathematics and in computer science from Aix-Marseille University, then in 2003, a PhD from Avignon University. She has been a teacher-researcher in mathematical optimization, since 2005 at the Computer Science Department of Mines Nantes, in charge of the GIPAD specialty (computer engineering for decision support), then at the Center for Applied Mathematics of Mines Paris-PSL since 2014. Attached to the fields of operational research and constraint programming, she serves to the scientific societies (ROADEF, Association for Constraint Programming, PGMO) and program committees.

Her work focuses on solving complex decision-making problems, combining discrete decisions and non-linear dynamics, and aims to improve industrial processes and inform public policies in the context of transition: planning, sizing, operating systems to minimize environmental impact, based on rich declarative models taking into account the constraints and externalities of decision-making, and through modular, efficient and flexible algorithms. Her methodological contributions mainly concern the hybridization and coordination of mathematical programming techniques, graphs, automata, constraint programming, machine learning, algebraic geometry. They are implemented, for example, in the design and operation of water, electrical and road networks; the management of data center resources; the electrification of heavy vehicle fleets; and staff planning.

Publication(s)

Teaching

Mathematical Optimization for the Transition

Course Director

The course is divided into two thematic parts, each alternating between lectures, tutorials (modeling exercises, convergence proofs), and lab sessions (implementation of the Gurobi solver): A. Integer Optimization for Discrete Problems 1. Review of Linear Optimization (Modeling Exercises, Geometry and Algebra, Duality and Optimality, Simplex Algorithm) 2. Integer Programming Modeling (Discrete Decisions, Logical Conditions, Nonlinear Functions, Ideal Formulations, Case Studies for the Transition) 3. Linear Integer Programming Algorithms (Cut Generation, Branch-and-Bound, Reformulation, and Decompositions) 4. Implementation (Branch-and-Cut, Modern Solvers and Parameterization, Case Study: Hybrid Electricity Generation) B. Equilibrium Problems from an Optimization Perspective 1. Review of Nonlinear Optimization (Problem Definition, Definition of Local and Global Solutions, Constraint Qualifications, First-Order Optimality Conditions, Second-Order Optimality Conditions, Duality) 2. Introduction to Complementarity Problems (Linear Complementarity Problems, Mixed Linear Complementarity Problems, Connection Between Optimization Problems and Complementarity Problems, Numerical Algorithms for Solving Complementarity Problems) 3. Optimization Problems Constrained by Complementarity Problems (Definition of Equilibrium Problems, Connection to Nash Equilibrium Problems, Transformation of a Nonconvex Quadratic Simplex Problem into a Mathematical Programming Problem with Complementarity Constraints) 4. Examples (Energy markets, International agreements on climate change adaptation) For each thematic section, course notes in slide format and the code needed for the lab will be made available to students. Equipment: personal computer with, as desired: a web browser and a GitHub or Google Drive account (for the code and to save the project in Google Colab) or a recent installation of Gurobi (several APIs available, including Python, C/C++, and Java) and MATLAB.

PhD supervision

  • 2026 Decarbonization of road freight transport: optimizing the formation and use of electric truck fleets AL KOSTIT Malak
  • 2025 Nuclear power program: technical-economic optimization and management of infrastructures and the nuclear fuel cycle GAGNEPAIN Albin
  • 2024 Mathematical programming with equilibrium constraints: models and algorithms for non-convex optimization MARTINS SASAKI Antonio
  • 2023 Hybrid combinatorial optimization and machine learning algorithms for energy-efficient water networks TAVAKOLI Amirhossein
  • 2022 Operational planning under joint uncertainties SYRTSEVA Ksenia
  • 2018 Integrating uncertainties in short-term operational planning JAVAL Paul
  • 2018 Optimal control and sizing of pumping stations in drinking water distribution networks BONVIN Gratien
  • 2017 Optimal control and energy management of an autonomous energy station through Robust Optimization N'GORAN Arnold
  • 2017 Integrated approach for prospective analysis of deep decarbonization of the European electricity system by 2050 in the face of climate variability SIGGINI Seyram
  • 2015 Robust optimization of control and sizing of an intelligent railway electric network at the scale of a station and its district HAVEL Aurélien
  • 2013 Dynamic multi-flows with resource synchronization constraints LIBEAUT Xavier
  • 2012 Bi-objective evaluation of the capacity of railway infrastructures using hybrid column generation MEREL Aurélien
  • 2011 Automata and constraint programming for personnel planning MENANA Julien