Review of artificial neural networks for gasoline, diesel and homogeneous charge compression ignition engine

Abstract

In automotive applications, artificial neural network (ANN) is now considered as a favorable prediction tool. Since it does not need an understanding of the system or its underlying physics, an ANN model can be beneficial especially when the system is too complicated, and it is too costly to model it using a simulation program. Therefore, using ANN to model an internal combustion engine has been a growing research area in the last decade. Despite its promising capabilities, the use of ANN for engine applications needs deeper examination and further improvement. Research in ANN may reach its maturity and be saturated if the same approach is applied repeatedly with the same network type, training algorithm and input–output parameters. This review article critically discusses recent application of ANN in ICE. The discussion does not only include its use in the conventional engine (gasoline and diesel engine), but it also covers the ANN application in advanced combustion technology i.e., homogeneous charge compression ignition (HCCI) engine. Overall, ANN has been successfully applied and it now becomes an indispensable tool to rapidly predict engine performance, combustion and emission characteristics. Practical implications and recommendations for future studies are presented at the end of this review.

Keywords

• Artificial neural network
• Gasoline engine
• Diesel engine
• Homogeneous charge compression ignition engine
• Internal combustion engine
• Optimization