Block sparse linear models for learning structured dynamical systems in aeronautics

Cédric Rommel 1, 2, 3, * Joseph Frédéric Bonnans 1, 3 Baptiste Gregorutti 2 Pierre Martinon 1, 3
* Corresponding author
3 Commands - Control, Optimization, Models, Methods and Applications for Nonlinear Dynamical Systems
CMAP - Centre de Mathématiques Appliquées - Ecole Polytechnique, Inria Saclay - Ile de France, UMA - Unité de Mathématiques Appliquées
Abstract : This paper addresses an aircraft dynamical system identification problem, with the goal of using the learned models for trajectory optimization purposes. Our approach is based on multi-task regression. We present in this setting a new class of estimators that we call Block sparse Lasso, which conserves a certain structure between the tasks and some groups of variables, while promoting sparsity within these groups. An implementation leading to consistent feature selection is suggested, allowing to obtain accurate models, which are suitable for trajectory optimization. An additional regularizer is also proposed to help in recovering hidden representations of the initial dynamical system. We illustrate our method with numerical results based on real flight data from 25 medium haul aircraft, totaling 8 million observations.
Complete list of metadatas

Cited literature [16 references]  Display  Hide  Download

https://hal.inria.fr/hal-01816400
Contributor : Cédric Rommel <>
Submitted on : Friday, June 15, 2018 - 11:32:27 AM
Last modification on : Friday, April 19, 2019 - 4:55:31 PM
Long-term archiving on : Monday, September 17, 2018 - 11:06:14 AM

File

paper.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01816400, version 1

Citation

Cédric Rommel, Joseph Frédéric Bonnans, Baptiste Gregorutti, Pierre Martinon. Block sparse linear models for learning structured dynamical systems in aeronautics. 2018. ⟨hal-01816400⟩

Share

Metrics

Record views

204

Files downloads

83