An Adaptive Neuro-fuzzy Modelling of Diesel Spray Penetration

The aim of this study was to demonstrate the effectiveness of an adaptive neuro-fuzzy inference system (ANFIS) for the prediction of diesel spray penetration length in an internal combustion engine. The technique involved extraction of necessary representative features from a collection of raw image data. An ANFIS was used to train the fuzzy inference system (FIS) and model the penetration length under different engine operating parameters, for example: in-cylinder pressure and temperature. The data obtained experimentally from the engine test rig was pre-processed using curve-fitting and averaging techniques. The devised mapping was compared with the experimental results and reasonable prediction was achieved. The results indicate that ANFIS can be used for modelling in-cylinder fuel spray behaviour as well as other operating parameters, potentially achieving very satisfactory results.