Neural network classification of diesel spray images

Abstract. This paper describes an evaluation of a neural network technique for modelling fuel spray penetration in the cylinder of a diesel internal combustion engine. The model was implemented using a multi-layer perceptron neural network. Two engine operating parameters were used as inputs to the model, namely injection pressure and in-cylinder pressure. Spray penetration length were modelled on the basis of these two inputs. The model was validated using test data that had not been used during training, and it was shown that semiautomated classification of complex diesel spray data is possible. The work lays the foundations for the establishment of an improved neural network paradigm for totally automatic, fast, accurate analysis of such complex data, thus saving many man-hours of tedious manual data analysis.