Prediction of performance parameters using artificial neural network and regression models in a bio-diesel engine using Honge methyl ester

In this paper, ANN modelling has been investigated for predicting the performance parameters of a diesel engine, run using Honge methyl ester. Multi-layer perceptron (MLP) has been used as the ANN model. The load acting on the engine, compression ratio, blend percentage, and the injection pressure were used as the input parameters while BTE, BSEC, and T_exh were used as the output parameters. The ANN-predicted results matched well with the experimental data over a wide range of operating parameters. The experimental and ANN predicted values were compared with the output from the regression models, which were developed for different output parameters. The significance of the input parameters and their influence on the outputs was evaluated using ANOVA. It was found that ANN models gave better prediction results compared to regression models.