Modelling and gasses emissions prediction for a turbo-charged diesel engine using artificial neural networks

Vehicles are one of the most important sources of environmental pollution, which make amajor concern in the public health and environment issues. When fuel is burned in the combustion engines, waste emitted from the car exhaust. The most serious of these waste are carbon monoxide and nitrogen oxides [1]. Techniques such as EGR and WI were used to reduce the ratio of gases emissions. This paper presents the design of artificial neural networks to estimate the amount of emitted gases in three combustion systems. The first uses Water Injection (WI) technique, the second uses Exhaust Gas Re-circulation (EGR) technique and the third uses a combine technique of them. The results obtained from the neural networks were compared with the measurements in the practical experiment. The test results showed that the neural network provide a good identification for the three systems.