查看文章

In this study statistical analysis methods were used for optimizing a spark ignition engine fueled by NG and hydrogen mixtures. Firstly designs of experiment and range analysis of the results have been carried out in order to improve the efficiency of experiments and reduce the workload. And then, a flexible model of this kind of engine that is catered to multidimensional optimization has been built. After that, the genetic algorithm is used to optimize the model. Finally the optimum control parameters of this operated point are determined to be hydrogen fraction 30–40%, excess air ratio 1.45–1.6 and ignition timing 20–22° BTDC at 1200r/min, 0.4MPa. The comparison of the optimized results and the original CNG performance showed that CH₄, CO, NO_x, and BSFC decrease by 70%, 83.57%, 93%, and 5%, respectively. This proved that the combination of artificial neural network and genetic algorithm is an effective way …