基于主元特征提取的SOFM网络实现飞行数据智能处理

The Intelligent Approach to Flight Data Processing Based on SOFM Neural Network Derived From Principle Character Drilling

为挖掘飞行参数中系统状态信息,用主元特征提取方法按判读需要降低参数的维数,对每帧数据按隶属度获得特征向量后输入SOFM网络,利用该网络的自组织分类功能进行系统状态的有效识别。最后对某型发动机工作状态进行识别,结果证明方法准确、可靠。

In order to dig the system state information from flight data in this paper,we use PCD to depree the dimension of system parameter according to the request of decision.The character vector of flight data pattern is distilled every frame and input to the SOFM which has the function of self-organization sorting to identify the system state effectively.At last,this way is proved to be nicety and credibility through the example to identify the engine sate of some kind.