摘要
以HL495Q型电喷汽油机为研究对象,分析了汽油机空燃比的数学模型,提出了一种基于Elman神经网络的过渡工况空燃比辨识方法。试验结果表明,Elman神经网络空燃比模型具有简单的网络结构,能高精度地逼近车用汽油机空燃比的实际动态过程,模型的平均相对误差小于1%,优于前馈BP神经网络模型的辨识结果。建立的Elman神经网络空燃比模型能改善过渡工况空燃比控制精度,提高排放性能。
Using HL495Q engine as research object, mathematical model of air fuel ratio of gasoline engine was analyzed and a method of indentifying air fuel ratio of gasoline engine during transient conditions based on Elman neural network was provided. Experiment results show that the air fuel ratio model of gasoline engine based on Elman neural network has simple structure and can accurately approximate the actual air fuel ratio transient process and average relative error is less than 1 %, and is prior to the air fuel ratio model based on feedforward BP neural network. The air fuel ratio model based on Elman neural network can improve air fuel ratio control accuracy and emission performance of gasoline engine during transient conditions.
出处
《中国公路学报》
EI
CAS
CSCD
北大核心
2006年第6期113-117,共5页
China Journal of Highway and Transport
基金
国家自然科学基金项目(50276005)
关键词
汽车工程
汽油机
ELMAN神经网络
空燃比
过渡工况
辨识
automotive engineering
gasoline engine
Elman neural network
air fuel ratio
transient condition
identification
