Accuracy of Friction Estimation during Driving

Hans-Ulrich Kobialka, Cornelia Lex

Research output: Chapter in Book/Report/Conference proceeding › Conference paper › peer-review

Abstract

Accuracy of online friction estimation depends on the ability of the sensors to capture information about the current interaction between road and tire. Sensors have differ-ent characteristics and limitations, so depending on the situation their contribution varies. In this work we investigated the construction of a model that maps a driving situation (represented as sensor data time series) to the accuracy of friction estimation that can be expected for this particular situation. To train such a model from data, we used „Echo State Networks“, a method for constructing and training large Recurrent Neural Networks.
Original languageEnglish
Title of host publicationConference Design of Experiments (DOE) in Engine Development
Place of PublicationRenningen
Publisherexpert verlag GmbH
Pages190-197
ISBN (Print)978-3-8169-3217-8
Publication statusPublished - 2013
EventConference Design of Experiments (DOE) in Engine Development - Berlin, Germany
Duration: 18 Jun 2013 → 19 Jun 2013

Publication series

NameDesign of Experiments (DoE) in Engine Development
PublisherExpert Verlag

Conference

ConferenceConference Design of Experiments (DOE) in Engine Development
Country/TerritoryGermany
CityBerlin
Period18/06/13 → 19/06/13

Fields of Expertise

  • Mobility & Production

Treatment code (Nähere Zuordnung)

  • Application
  • Theoretical

Cite this