摘要
油膜动态效应的存在对PFI汽油机油瞬态空燃比的精确控制具有较大影响,而油膜参数又是油膜动态效应中最关键的参数。为提高油膜参数的辨识精度,提出了一种组合的混沌粒子群优化算法(CPSO),并在Simulink中建立了基于CPSO-RBF神经网络的汽油机瞬态工况油膜参数辨识模型。将辨识得到的油膜参数以与BP神经网络辨识及最小二乘辨识得到的结果进行对比,结果表明:CPSO-RBF神经网络辨识方法能对油膜参数进行有效辨识,具有更强的非线性辨识能力和更高的辨识精度。
The existence of oil film dynamic effect has a great influence on the accurate control of the tran-sient air-fuel ratio of PFI gasoline engine oil, and the oil film parameter is the most critical parameter inthe dynamic effect of the oil film. To improve identification accuracy of oil film parameter, proposed chaot-ic particle swarm optimization algorithm is a combination of(CPSO), and the establishment of gasoline en-gine in transient condition of oil film parameter identification model based on CPSO-RBF neural networkin Simulink. Oil film parameter theidentified obtained by BP neural network identification and least squaresidentification results were compared, the results show that the identification method of CPSO-RBF neuralnetwork can effectively identify the oil film parameters, nonlinear identification ability is stronger and high-er degree of identification precision.
出处
《河南科技》
2017年第21期129-132,共4页
Henan Science and Technology
基金
湖南省自然科学基金资助项目(2016JJ2003)
关键词
油膜参数
瞬态空燃比
CPSO优化算法
最小二乘辨识
fuel film parameters
transient air-fuel ratio
CPSO algorithm
least squares identification