Skip to Main content Skip to Navigation
Conference papers

Combining experimental data and physical simulation models in Support Vector learning

Abstract : This paper considers modeling the in-cylinder residual gas fraction in Spark Ignition (SI) engine with Variable Camshaft Timing (VCT) based on a limited amount of experimental data and a simulator built from prior knowledge. The problem of how to best incorporate the data provided by the simulator, possibly biased, into the learning of the model is addressed. This problem, although particular, is very representative of numerous situations met in engine control, and more generally in engineering, where complex models, more or less accurate, exist and where the experimental data which can be used for calibration are difficult or expensive to obtain. The first proposed method applies a different loss function on the simulation data allowing for a certain level of inaccuracy. The second method constrains the derivatives of the model to be determined to fit to the derivatives of a prior model previously estimated on the simulation data. Finally, a third method considers the combination of these two forms of prior knowledge. These approaches are implemented in the linear programming support vector regression (LP-SVR) framework by the addition of constraints linear in the parameters to the optimization problem. Promising results are obtained on the application.
Document type :
Conference papers
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-00152249
Contributor : Gérard Bloch <>
Submitted on : Tuesday, June 12, 2007 - 11:37:58 AM
Last modification on : Friday, March 15, 2019 - 4:48:11 PM
Long-term archiving on: : Thursday, April 8, 2010 - 7:04:32 PM

File

BlochLauerColinChamaillardEANN...
Files produced by the author(s)

Identifiers

  • HAL Id : hal-00152249, version 1

Collections

Citation

Gérard Bloch, Fabien Lauer, Guillaume Colin, Yann Chamaillard. Combining experimental data and physical simulation models in Support Vector learning. 10th International Conference on Engineering Applications of Neural Networks, EANN 2007, Aug 2007, Thessaloniki, Greece. pp.284-295. ⟨hal-00152249⟩

Share

Metrics

Record views

202

Files downloads

162