RBF神经网络软测量技术在汽油机CO排放中的应用

RBF Neural Networks Technology in Emission of Gasoline Engine

以汽油机排放污染物CO难于进行实时控制测量为例,讨论了神经网络技术在汽油机中应用的前景,提出了神经网络软测量的概念,建立了基于汽油机排放污染物CO的神经网络控制模型,并应用MATLAB软件编程对该神经网络进行学习训练。结果表明:该模型可在没有专用的汽车排放仪器时进行CO排放值的测量,也可用于CO的实时控制。

It is difficult to immediately measure the CO emission of gasoline engine. The prospect of applying for neural network technique in the gasoline engine was discussed in this paper. The concept of neural network soft-measurement is advanced and the neural network control model of CO emission in the gasoline engine was also established. The neural network model was trained based on MATLAB software. The training results show that this model can measure the CO emission without the instruments, and can immediately control the CO emission.