中文摘要：

针对电控柴油机控制参数虚拟优化中传统优化算法存在的一些问题,利用一种全局智能寻优方法——微种群遗传算法,并结合改进的发动机三维数值模拟程序KIVA3V,建立起一个发动机多参数优化软件平台.在该软件平台上,对带有BUMP燃烧室的电控柴油机的喷油时刻和喷油压力进行优化.结果显示,在平均有效压力为0.79MPa时,优化后的发动机排放与原机相比,NOx由4.11 g/（kW.h）降低到2.51 g/（kW.h）,碳烟由0.267 g/（kW.h）降低到0.240 g/（kW.h）,优化结果与优化试验数据基本吻合,说明微种群遗传算法非常适合发动机多参数优化问题的求解,优化结果可为实际燃烧过程的优化控制提供参考方案.

英文摘要：

There exist some difficulties when traditional optimization algorithms are used to solve parameter optimization for electronically controlled diesel engines. The developed micro-genetic algorithm is integrated into the modified KIVA-3V code for optimizing engine parameters in this study. The optimization study of start-of-injection timing and injection pressure is performed for the electronically controlled diesel engine with a bump design combustion chamber. The results show that, at the indicated mean effective pressure of 0.79 MPa, the optimized parameters result in significantly lower engine emissions compared with the baseline engine, NOx decreasing from 4.11 g/（ kW · h） to 2.51 g/（ kW · h） and soot from 0. 267 g/（ kW · h） to 0.240 g/（ kW · h）, which demonstrates that the micro-genetic algorithm can be suited to optimize the parameters of electronically controlled diesel engines.