Optimization of homogeneous charge compression ignition with genetic algorithms

J. Y. Chen*, Robert Dibble, J. Kolbu, R. Homma

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

27 Scopus citations

Abstract

Homogeneous charge compression ignition (HCCI) engines promise high efficiency and low emissions, and thus these engines hold great potential for reducing pollution in electric power generation, trucks, marine vehicles, locomotives, and automobiles. Controlling the time of the HCCI combustion event remains a major technical difficulty yet to be overcome. An integration of genetic algorithms (GAs) with well-mixed reactor simulations is developed for better understanding HCCI combustion and for guiding the development of optimal controls. With GAs, the effects of engine intake charge on engine performance are explored with three fuels: methane, propane, and acetylene. Simulation results are compared to available experimental data showing that model predictions are consistent with known trends. As the first application, GAs are used for searching the optimal intake charge conditions of HCCI combustion for power generation with methane. Results suggest that the use of intake charges with high equivalence ratio and with large amounts of exhaust gas recirculation is optimal. Subsequently, parameter optimization for intake conditions with the stoichiometric mixture at various power demand is explored. The results are analyzed and insights on the optimal managing schemes are revealed.

Original languageEnglish (US)
Pages (from-to)373-392
Number of pages20
JournalCombustion science and technology
Volume175
Issue number2
DOIs
StatePublished - Feb 1 2003

Keywords

  • Genetic algorithms
  • HCCI combustion

ASJC Scopus subject areas

  • Chemistry(all)
  • Chemical Engineering(all)
  • Fuel Technology
  • Energy Engineering and Power Technology
  • Physics and Astronomy(all)

Fingerprint Dive into the research topics of 'Optimization of homogeneous charge compression ignition with genetic algorithms'. Together they form a unique fingerprint.

Cite this