基于遗传算法的柴油PCCI发动机喷射参数的优化  

Optimization of Injection Parameters for Diesel PCCI Combustion Based on Genetic Algorithm

作　　者：陈石[1] 林春伟[1] 贾明[1] 庞斌[1] 解茂昭[1] 

机构地区：[1]大连理工大学能源与动力学院,辽宁大连116024

出　　处：《内燃机学报》2011年第5期414-421,共8页Transactions of Csice

基　　金：国家基础研究重点规划子课题资助项目(2007CB210002);国家自然科学基金资助项目(50806008)

摘　　要：利用遗传算法优化预混压燃(PCCI)柴油发动机的喷油压力、喷射夹角和喷射定时3个喷射参数的匹配关系,以达到氮氧化物和碳烟排放及指示燃油消耗率的最佳值.计算中将非支配排序遗传算法(NSGA2)、KIVA-3V和主从式并行算法(Master-Slave)耦合成KIVA3V-NSGA2-MS程序,优化PCCI燃烧以搜寻最优的匹配喷射参数,同时利用缸内当量比分布图及喷雾定位示意图详细分析每个参数对燃烧和排放的影响.结果发现,与基础试验值相比,优化得到的一个Pareto最优解,NOx排放降低17.34%,碳烟排放降低58.23%.当喷射夹角为80°、喷油压力为149.7,MPa、喷射定时在27°CA,BTDC之前时,缸内完全是预混燃烧,高温燃烧始点均在上止点前11°CA附近.在PCCI燃烧模式下,由于喷油始点的提前,提高喷射压力,对燃烧和排放的影响不明显.研究燃油和空气预混程度以及燃烧排放值,需要综合考虑喷射定时和喷射夹角因素.Genetic algorithm （GA）is used to optimize the injection parameters including the start of injection （SOl）, injection pressure and spray angle for diesel premixed charge compression ignition （PCCI） combustion. A non-dominated sorting genetic algorithm （NSGA2）was coupled with the computational fluid dynamics （CFD）code KIVA-3V, as well as with a master-slave parallel algorithm to conduct the multi-objective optimization to get low emissions and good fuel economy. The effect of each parameter on combustion and emissions was investigated based on the distributions of temperature and equivalence ratio, and the spray impinging on cylinder. Study reveals that premixed combustion dominates PCCI combustion process when SOI advances before 27° CA BTDC for spray angle of 80° and 149.7 MPa injection pressure. In the PCCI strategy, injection pressure does not have an obvious effect on combustion and emissions. Spray angle and emissions should consider in determining the SOI.

关 键 词：预混压燃 遗传算法 燃烧优化 排放 

分 类 号：TK401[动力工程及工程热物理—动力机械及工程]