The Multi-Parameter Optimization of Injections on Double-Layer Diesel Engines Based on Genetic Algorithm

In this paper, a coupling model of the NSGA-II genetic algorithm and KIVA-3 V program is developed to perform a multi-objective intelligent optimization study of the injection parameters in a diesel engine with double-layer diverging combustion (DLDC) chamber. Four optimization parameters, involving the first injection timing, the first fuel split ratio, the injection timing interval, and the exhaust gas recirculation (EGR) ratio, are optimized after comparing the pollutant emissions and fuel consumption characteristics. The computation results reveal that both soot and fuel consumption function as the optimization objectives, proving a trade-off relationship with the NOx emission. The analysis of the response surface demonstrates that the application of the multiple-injection strategies is beneficial for enhancing the performance of diesel engines. A large first fuel split ratio and a low EGR rate can minimize both ISFC (indicated specific fuel consumption) and soot emission concurrently. A high EGR rate can significantly reduce NOx emission. In addition, the distribution of the final Pareto font shows that 66.7 % percent of the first fuel split ratio is above 0.56, and only two solutions (8.3 %) have values below 0.35. Furthermore, three representative points of the first fuel split ratio are selected for combustion analysis, which are 0.05, 0.35, and 0.88. As the first fuel split ratio rises, the high-temperature areas in the cylinder are gradually enlarged, and the NOx emission is increased while the soot emission and ISFC are further reduced.