Publications

Structured Representation Learning for Behavior Cloning: How can we learn to safely control a nuclear power plant?

with Alain Grossetête & Nicolas Vayatis

Structuring the latent space by timescale makes behavior cloning of an NMPC expert for nuclear load-following more accurate and more often feasible — and, used to warm-start the NMPC, it recovers full feasibility while cutting computation time by 15%.

• Imitation Learning
• Machine Learning
• Model Predictive Control

Accelerating Full-Scale Nonlinear Model Predictive Control via Surrogate Dynamics Optimization

with Guillaume Dupré, Alain Grossetête & Nicolas Vayatis

In long-horizon nonlinear MPC, we show that learned surrogate dynamics can significantly accelerate computation while preserving safety.

• Dynamical Systems
• Machine Learning
• Model Predictive Control

Enhancing Nuclear Reactor Core Simulation through Data-Based Surrogate Models

with Alain Grossetête & Nicolas Vayatis

This paper develops a ML surrogate simulation scheme for fast integration of Differential Algebraic Equations modeling a nuclear reactor core.

• Machine Learning
• Dynamical Systems

LIPO+: Frugal Global Optimization for Lipschitz Functions

with Gaëtan Serré, Sophia Chirrane, Argyris Kalogeratos & Nicolas Vayatis

This paper develops two major improvements on the global optimization method LIPO from Malherbe & Vayatis, 2017: an empirical stopping criterion and a decaying exploration rate.

• Machine Learning
• Optimization

Overview of AdaBoost: Reconciling its views to better understand its dynamics

Unifying the views of AdaBoost, in order to better understand its dynamics.

• Machine Learning
• Dynamical Systems

A novel warm-start method for Optimal Control of Nuclear Power Plants

with Alain Grossetête & Nicolas Vayatis

Using a ML surrogate simulation scheme for fast integration of Differential Algebraic Equations, we design a warm-start method for NMPC applied to the load-following of Nuclear Power Plants.