arXiv subjects

Florent Di Meglio

Publications and source records attributed to Florent Di Meglio.

10 recordsLinked to original sources

arXiv2026

Domain Generalization for Time Series: Enhancing Drilling Regression Models for Stick-Slip Index Prediction

This paper provides a comprehensive comparison of domain generalization techniques applied to time series data within a drilling context, focusing on the prediction of a continuous Stick-Slip Index (SSI), a critical metric for assessing torsional downhole vibrations at the drill bit. The study aims to develop a robust regression model that can generalize across domains by training on 60 second labeled sequences of 1 Hz surface drilling data to predict the SSI. The model is tested in wells that are different from those used during training. To fine-tune the model architecture, a grid search approach is employed to optimize key hyperparameters. A comparative analysis of the Adversarial Domain Generalization (ADG), Invariant Risk Minimization (IRM) and baseline models is presented, along with an evaluation of the effectiveness of transfer learning (TL) in improving model performance. The ADG and IRM models achieve performance improvements of 10% and 8%, respectively, over the baseline model. Most importantly, severe events are detected 60% of the time, against 20% for the baseline model. Overall, the results indicate that both ADG and IRM models surpass the baseline, with the ADG model exhibiting a slight advantage over the IRM model. Additionally, applying TL to a pre-trained model further improves performance. Our findings demonstrate the potential of domain generalization approaches in drilling applications, with ADG emerging as the most effective approach.

cs.LG↗

arXiv2024

Data-Driven Observability Analysis for Nonlinear Stochastic Systems

Distinguishability and, by extension, observability are key properties of dynamical systems. Establishing these properties is challenging, especially when no analytical model is available and they are to be inferred directly from measurement data. The presence of noise further complicates this analysis, as standard notions of distinguishability are tailored to deterministic systems. We build on distributional distinguishability, which extends the deterministic notion by comparing distributions of outputs of stochastic systems. We first show that both concepts are equivalent for a class of systems that includes linear systems. We then present a method to assess and quantify distributional distinguishability from output data. Specifically, our quantification measures how much data is required to tell apart two initial states, inducing a continuous spectrum of distinguishability. We propose a statistical test to determine a threshold above which two states can be considered distinguishable with high confidence. We illustrate these tools by computing distinguishability maps over the state space in simulation, then leverage the test to compare sensor configurations on hardware.

eess.SY↗

arXiv2023

Enabling safe walking rehabilitation on the exoskeleton Atalante: experimental results

This paper exposes a control architecture enabling rehabilitation of walking impaired patients with the lower-limb exoskeleton Atalante. Atalante's control system is modified to allow the patient to contribute to the walking motion through their efforts. Only the swing leg degree of freedom along the nominal path is relaxed. An online trajectory optimization checks that the muscle forces do not jeopardize stability. The optimization generates reference trajectories that satisfy several key constraints from the current point to the end of the step. One of the constraints requires that the center or pressure remains inside the support polygon, which ensures that the support leg subsystem successfully tracks the reference trajectory. As a result of the presented works, the robot provides a non-zero force in the direction of motion only when required, helping the patient go fast enough to maintain balance (or preventing him from going too fast). Experimental results are reported. They illustrate that variations of $\pm$50% of the duration of the step can be achieved in response to the patient's efforts and that many steps are achieved without falling. A video of the experiments can be viewed at https://youtu.be/_1A-2nLy5ZE

cs.RO↗

arXiv2023

Towards gain tuning for numerical KKL observers

This paper presents a first step towards tuning observers for general nonlinear systems. Relying on recent results around Kazantzis-Kravaris/Luenberger (KKL) observers, we propose an empirical criterion to guide the calibration of the observer, by trading off transient performance and sensitivity to measurement noise. We parametrize the gain matrix and evaluate this criterion over a family of observers for different parameter values. We then use neural networks to learn the mapping between the observer and the nonlinear system, and present a novel method to sample the state-space efficiently for nonlinear regression. We illustrate the merits of this approach in numerical simulations.

eess.SY↗

arXiv2023

Recognition Models to Learn Dynamics from Partial Observations with Neural ODEs

Identifying dynamical systems from experimental data is a notably difficult task. Prior knowledge generally helps, but the extent of this knowledge varies with the application, and customized models are often needed. Neural ordinary differential equations can be written as a flexible framework for system identification and can incorporate a broad spectrum of physical insight, giving physical interpretability to the resulting latent space. In the case of partial observations, however, the data points cannot directly be mapped to the latent state of the ODE. Hence, we propose to design recognition models, in particular inspired by nonlinear observer theory, to link the partial observations to the latent state. We demonstrate the performance of the proposed approach on numerical simulations and on an experimental dataset from a robotic exoskeleton.

eess.SY↗

arXiv2022

Fast replanning of a lower-limb exoskeleton trajectories for rehabilitation

The paper addresses the rehabilitation of disabled patients using a lower-limb fully-actuated exoskeleton. We propose a novel numerical method to replan the current step without jeopardizing stability. Stability is evaluated in the light of a simple linear time-invariant surrogate model. The method's core is the analysis of an input-constrained optimal control problem with state specified at an unspecified terminal time. A detailed study of the extremals given by Pontryagin Maximum Principle is sufficient to characterize its feasibility. This allows a fast replanning strategy. The efficiency of the numerical algorithm (resolution time below 1 ms) yields responsiveness to the patient's request. Realistic simulations on a full-body model of the patient-exoskeleton system stress that cases of practical interest for physiotherapists are well-addressed.

math.OC↗

arXiv2017

Robust output regulation of 2 x 2 hyperbolic systems part I: Control law and Input-to-State Stability

We consider the problem of output feedback regulationfor a linear first-order hyperbolic system with collocatedinput and output in presence of a general class of disturbancesand noise. The proposed control law is designed through abackstepping approach incorporating an integral action. Toensure robustness to delays, the controller only cancels partof the boundary reflection by means of a tunable parameter.This also enables a trade-off between disturbance and noisesensitivity.We show that the boundary condition of the obtainedtarget system can be transformed into a Neutral DifferentialEquation (NDE) and that this latter system is Input-to-StateStable (ISS). This proves the boundedness of the controlledoutput for the target system. This extends previous worksconsidering an integral action for this kind of system [16], andconstitutes an important step towards practical implementationof such controllers. Applications and practical considerations,in particular regarding the system's sensitivity functions arederived in a companion paper.

math.AP↗

arXiv2015

Minimum time control of heterodirectional linear coupled hyperbolic PDEs

We solve the problem of stabilization of a class of linear first-order hyperbolic systems featuring n rightward convecting transport PDEs and m leftward convecting transport PDEs. Using the backstepping approach yields solutions to stabilization in minimal time and observer based output feedback.

math.AP↗

arXiv2015

Boundary exponential stabilization of 1-D inhomogeneous quasilinear hyperbolic systems

This paper deals with the problem of boundary stabilization of first-order n\times n inhomogeneous quasilinear hyperbolic systems. A backstepping method is developed. The main result supplements the previous works on how to design multi-boundary feedback controllers to realize exponential stability of the original nonlinear system in the spatial H^2 sense.

math.OC↗

arXiv2015

Control of Homodirectional and General Heterodirectional Linear Coupled Hyperbolic PDEs

Research on stabilization of coupled hyperbolic PDEs has been dominated by the focus on pairs of counter-convecting ("heterodirectional") transport PDEs with distributed local coupling and with controls at one or both boundaries. A recent extension allows stabilization using only one control for a system containing an arbitrary number of coupled transport PDEs that convect at different speeds against the direction of the PDE whose boundary is actuated. In this paper we present a solution to the fully general case, in which the number of PDEs in either direction is arbitrary, and where actuation is applied on only one boundary (to all the PDEs that convect downstream from that boundary). To solve this general problem, we solve, as a special case, the problem of control of coupled "homodirectional" hyperbolic linear PDEs, where multiple transport PDEs convect in the same direction with arbitrary local coupling. Our approach is based on PDE backstepping and yields solutions to stabilization, by both full-state and observer-based output feedback, trajectory planning, and trajectory tracking problems.

math.OC↗