Summary

PhD candidate in Fluid Mechanics. Research: ML + EIT for electric pump energy optimization

Education

• PhD in Fluid Mechanics
  Aix-Marseille University, Laboratory of Mechanics and Acoustics (LMA) & FLUIIDD
• Mobile Autonomous Systems

  2024
  Université Paris-Saclay
• Engineering & Master 2 in Control Systems

  2023
  École Nationale Polytechnique
• Classes préparatoires en science et technologie

  2020
  École Nationale Polytechnique

Work Experience

• PhD Candidate (regional-funded thesis)

  2025 -
  Optimizing electric pump energy consumption using machine learning and electrical impedance tomography.
• Research internship in Artificial Intelligence

  2024 - 2024
  Design of an autonomous driving algorithm using large language model inspired control approaches on the CARLA simulator.
• Research internship in Control Theory

  2023 - 2023
  Arterial blood flow estimation using modulating function methods.

Skills

Publications

• B-spline-based Modulating Function Method for Arterial Blood Flow's Estimations

  2024
  IFAC-PapersOnLine, Volume 58, Issue 24, Pages 573-578, 2024

  This paper considers the estimation problem of arterial blood flow by using a fractional model of arterial hemodynamics. This model has been introduced as an enhanced version of the conventional integer-order Windkessel model. This advanced model incorporates a fractional-order capacitor to characterize the intricate and the frequency-dependent arterial compliance. A recent algorithm that consists of modulating functions method combined with an iterative Newton approach has been proposed to solve this estimation problem. However, the method’s performance was highly dependent on the convergence of the Newton method. In addition, it was affected by the choice of the basis functions. In this paper, we present an enhanced version of this two-stage algorithm, which leverages modulating functions, genetic algorithms, and B-spline basis functions for the simultaneous estimation of both the model’s input (the blood flow) and the fractional differentiation order within a finite time frame. To validate the approach, in silico human data is employed, and its performance is compared with that of the former method. The outcomes demonstrate significant potential for accurate calibration of the fractional model and estimation of the blood flow.

  View Publication

Teaching

• Arduino Programming

  05-0
  Aix-Marseille University, IUT of Sciences

  Role: L3 GMP

Languages

• French

  Fluent (TCF C1)
• English

  Fluent (CEFR C1+)
• Arabic

  Native
• Italian

  Beginner (A1)