Genetic algorithm and artificial neural network for engine optimisation of efficiency and NOx emission

Genetic algorithm (GA) and neural network analysis are used to predict the effects of design and operational parameters on engine efficiency and NOx emissions of a natural gas engine. A computer program to calculate the amount of NOx emissions based on a reaction kinetic model is developed. The validity of this program is verified by measurements from a turbocharged, lean-burn, natural gas engine. Using the results from this program, the effects of operational and design parameters of the engine were investigated. Then a wide range of engine parameters are optimised using a simple GA regarding both efficiency and NOx emissions. Because of the large computation requirements especially for NOx level determination, an artificial neural network model based on results of these investigations is used to predict the engine efficiency and NOx emissions. The results show an increase in efficiency as well as the amount of NOx emissions being kept under the constraint value of 250 mg/Nm3 for stationary engines.