RBF网络在柴油机故障诊断领域中的应用与实现

Layout and Its Applications to Make a Diagnosis for Marine Diesel Engine by Using Radial Basis Transfer Function Network

阐述了神经网络的模型、算法和分类,主要分析径向基传递函数、径向基函数分布常数以及高斯函数的宽度系数,探讨MATLAB环境下采用径向基函数网络实现对柴油机故障诊断的方法,并与BP网络性能进行比较.表明RBF网络学习速度很快,适于在线实时监测与诊断.

A radial basis function network is a kind of forward feedback artificial network model,it has been gotten more and more applications in recent years.In this paper,the model,algorithm and classification of neural network are performed.Special emphasis has been placed on the transfer function,spread constant of radial basis function,and the width coefficient of Gauss function.Then in MATLAB environment, the approaches to fault diagnosis for marine diesel engine by using a RBF neural network have been implemented.As the same time, the result of performance is given compared with BP algorithms. It demonstrates that the learn rate of RBF neural network is rapid,and online monitoring and diagnosis is feasible.