摘要
在设计某型号国Ⅳ柴油机选择性催化还原系统SCR时,为了快速标定SCR控制器,采用空间填充试验设计得到130个优化工况点,使试验规模减小了2/3。基于排放特性试验数据库和催化器特性试验数据库,利用神经网络开发了目标柴油机排放模型和催化器模型。基于催化器模型,通过多目标优化计算得到尿素计量脉谱,ESC循环试验证明所标定的脉谱精度满足控制要求。
In order to quickly calibrate SCR controller,space-falling experiment design was applied to get 130 optimization operation points while developing SCR system for a type of diesel engine,which reduced the experimental workload for two-thirds.The target diesel engine emission model and SCR catalyst model were developed by using neural network based on diesel emission characteristic experimental database and catalyst characteristic experimental database.Multi-objective optimization arithmetic was applied to gain urea dosing map based on SCR catalyst model.ESC（Europe steady cycle） experiment validated that the precision of urea dosing map satisfied the requirement of SCR control.
出处
《农业机械学报》
EI
CAS
CSCD
北大核心
2012年第5期16-21,83,共7页
Transactions of the Chinese Society for Agricultural Machinery
基金
国家高技术研究发展计划(863计划)资助项目(2007AA06Z341)
