Data driven model for detecting misfiring combustion in low temperature combustion engine using Neuro-Fuzzy

The homogeneous charge compression ignition (HCCI) engines with ethanol fuel as a renewable fuel is a promising solution to some of the major challenges of combustion engines. Incomplete or misfiring combustion limits HCCI operation and damages the catalyst converter and exhaust systems. The experimental data of a 0.3-liter combustion engine was used for modeling and detecting misfiring combustion. Incomplete and misfiring combustion in HCCI engine was studied by fuzzy-neural network. There is a significant relationship between misfiring combustion and in-cylinder pressure variations at 0, 5, 10, 15 and 20 crankshafts. These experimental findings were used to design a fuzzy-neural network for misfiring incomplete combustion in a HCCI engine. This model has been tested on experimental data. The results showed that the fuzzy-neural network fault diagnostic model can detect incomplete and misfiring combustion in HCCI engine with ethanol fuel. In addition, the developed model was able to identify the transition success from the normal operating area and incomplete combustion.