Application of neural network for air-fuel ratio identification in spark ignition engine

  • Authors:
  • Samir Saraswati;Satish Chand

  • Affiliations:
  • Department of Mechanical Engineering, MNNIT, Allahabad 211004, India.;Department of Mechanical Engineering, MNNIT, Allahabad 211004, India

  • Venue:
  • International Journal of Computer Applications in Technology
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

In the present work, Recurrent Neural Network (RNN) is used forAir-Fuel Ratio (AFR) identification in Spark Ignition (SI) engine.AFR identification is difficult due to nonlinear and dynamicbehaviour of SI engines. Delays present in the engine dynamicslimits the performance of engine controller. Identifying AFR fewsteps in advance can help engine controller to take care of these.RNN is trained using data from engine simulations inMATLAB/SIMULINK© environment. Uncorrelated signals weregenerated for training and generalisation and it has been shownthat RNN can predict engine simulations with reasonably goodaccuracy. RNN discussed can also work as a virtual AFR sensor andit can very well replace costly AFR sensor used in SI engines.