基于多重结构神经网络的歼击机结构故障检测

提出了一种基于多重结构神经网络的故障检测方法。针对以歼击机为代表的非线性系统中存在的突发结构故障，构造了一个多重结构神经网络，在输入层对残差信号进行二进离散小波变换，提取其在多尺度下的细节系数作为故障特征向量，并将其输入到神经网络分类器进行相应的模式分类，之后再利用下一级的故障度辨识神经网络对故障的大小进行辨识。仿真结果表明，本文方法为歼击机组合结构故障的检测提供了有效的方法和途径。

Structural Fault Detection of Fighter Based on Multiple Neural Networks

A method based on multiple neural networks for structural fault is presented. Considering the structural faults in complex nonlinear system, such as fight aircraft, a multiple neural network is constructed. Residual signal is processed by discrete binary wavelet transformation at the input layer, and the detail coefficients are obtained under multi-resolution as fault character vectors, then these character vectors are sent into neural network classifier to complete fault pattern classification, finally neural network identifier gives the degree of fault. Simulational results reveal that the method is effective not only for single fault diagnosis, but also for multiple fault diagnosis.