基于PNN的音频检测硬度分类器研究

Sonic Testing for Classification of Hardness Basis on PNN

依据音频检测原理,针对某火炮零部件(凸轮轴)硬度与音频特征参数的非线性映射问题,提出了基于音频检测与概率神经网络(PNN)的分类方法,实现了不同零部件的硬度分类,运用PNN对音频参数与硬度值进行仿真,精度满足要求。

Aimed at unlinear mapped relation between the hardness ofartillery＇s parts and sonic parameters, based on sonic testing theory, this paper proposed classificatory method based on sonic testing and probabilistic neural network （ PNN ） , it realizes classification between different hardness, it also exerts PNN to simulate hardness with sonic parameters together. And its precision accords with experiments demand.