车用汽油机瞬态空燃比的混沌时序非线性组合辨识模型

Chaotic Sequence Nonlinear Combination Identification Model of Transient Air-fuel Ratio of Gasoline Engine

为提高汽油机瞬态空燃比的辨识精度,提出了混沌时序非线性组合辨识模型。采用2种单项辨识方法,包括最小二乘支持向量机(LS-SVM)及径向基函数(RBF)前向型神经网络,分别对瞬态空燃比时间序列进行建模与辨识。采用非线性组合方法利用BP神经网络对2种单项辨识方法的结果进行组合辨识,并与Elman神经网络模型及最小二乘辨识模型进行比较。结果表明:混沌时序非线性组合辨识模型的辨识精度优于Elman神经网络模型及最小二乘辨识模型,具有更强的非线性辨识能力,能提高瞬态空燃比的辨识精度,为空燃比反馈控制的成功实行提供了有力依据。

In order to improve the identification accuracy of transient air-fuel ratio of gasoline engine, a chaotic sequence nonlinear combination identification model was proposed. With two kinds of individual identification methods, including the least squares support vector machine （LS-SVM） and the radial basis function （RBF） forward neural network, the sequence of transient air-fuel ratio was modeled and identified respectively. With the nonlinear combination method and the BP neural network, the results of the two individual identification methods were combined and identified, which were compared with the identification methods of Elman neural network model and the recursion algorithm of the least-square. The results show that the identification accuracy of the chaotic sequence nonlinear combination identification model is superior to the Elman neural network model and the recursion algorithm of the least-square, with stronger nonlinear recognition ability, which can improve the identification accuracy of the transient air-fuel ratio and provide a strong basis for successful feedback control of air-fuel ratio.