Engine performance and emission analysis of LPG-SI engine with the aid of artificial neural network

Abstract

Artificial neural network (ANN) technique is used in this analysis to estimate the performance and emission concentration of liquefied petroleum gas (LPG) spark-ignition (SI) engine. The performance indicators include fuel consumption and brake thermal efficiency while the emission components are NO x, CO, CO2, O2, and unburned hydrocarbon (UHC). Data of engine body temperature and exhaust gas temperature are also simulated. The first part of this study involves experimental works where a single-cylinder four-stroke SI engine was modified to allow the intake of LPG and operated at variable loadings with constant speed. The experimental results show that LPG produces comparable performance at high loads while significant reduction takes place in NO x, CO, CO2, O2, and UHC concentrations. The second part of this study involves the use of back-propagation algorithm for the training of the ANN model. The result of the simulation reveals that ANN model is appropriate to estimate the engine performance and gas exhaust emissions with correlation coefficient ranging from 0.9 to 0.99 with low root mean-squared error and low mean relative error.