Data-driven algorithm to estimate friction in automobile engine

Alexander A. Stotsky

Research output: Contribution to journalJournal articleResearchpeer-review


Algorithms based on the oscillations of the engine angular rotational speed under fuel cutoff and no-load were proposed for estimation of the engine friction torque. The recursive algorithm to restore the periodic signal is used to calculate the amplitude of the engine speed signal at fuel cutoff. The values of the friction torque in the corresponding table entries are updated at acquiring new measurements of the friction moment. A new, data-driven algorithm for table adaptation on the basis of stepwise regression was developed and verified using the six-cylinder Volvo engine.
Original languageEnglish
JournalAutomation and Remote Control
Issue number6
Pages (from-to)1135-1146
Number of pages11
Publication statusPublished - Jun 2010

Cite this