基于遗传自适应RBFN的歼击机故障分类方法

Fault Classification Method for the Fighter Aircraft Based on Genetic a RBFN

在模式识别领域中,如何实现更高精度地分类一直是个核心问题。提出了将自适应RBF神经网络与遗传算法相结合的方法,其中自适应RBF神经网络通过对样本判断,自动实现对RBF网络添加新的隐层节点,或者将样本归于已存在的隐层节点所属的类。遗传算法用于寻找最优的网络宽度值。两者相结合最后确定一个隐层节点数与类别数相同的简省的网络。用歼击机故障数据进行仿真,比较结果表明此方法能实现更高精度的分类。

How to obtain a more accurate class separability has been a key question in the field of classification application. A method is proposed which combines an adaptive RBF（ radial basis function） neural network and a GA （genetic algorithm）. The adaptive RBFN is used to add new hidden layer neurons or to determine the certain class which the input vector belongs to. The GA is used to search for the best value for the parameter of RBFN by estimating the input vectors. The method which includes ARBFN and GA can select a parsimonious network architecture. Compared with other methods, the result shows that our method can achieve a more precise class separability.