Fuzzy Logic Based Prediction of Performance and Emission Parameters of a LPG-Diesel Dual Fuel Engine

Abstract

Use of dual fuel engine for reduction of emissions and improvement in performance is an ongoing research area. The focus in engine design and configuration has been towards reducing emissions to avoid environment pollution. LPG/diesel dual fuel engine have been found to reduce Nitrogen oxide (NOx) and smoke. Artificial Intelligence based methods like neural networks and fuzzy logic have been used to model and study engine performance and emission parameters. Their ability to learn from data, being fault tolerant, handle noisy and incomplete data and ability to deal with nonlinear problems can be useful in these applications. In this work, fuzzy logic has been used to model performance and emission parameters in Liquid Petroleum Gas (LPG)-diesel dual fuel engine. The performance parameters included brake specific energy consumption (BSEC) and brake thermal efficiency (BTE) and emission parameters included exhaust gas temperature (EGT) and smoke. Adaptive neuro fuzzy inference system (ANFIS), a hybrid technique involving fuzzy logic and artificial neural network (ANN) have been used to develop a model and its performance has been compared with conventional fuzzy logic based model. It was found that ANFIS outperformed conventional fuzzy logic model based on R2 value and prediction accuracy on test data.