光纤围栏扰动事件模式识别研究

Research on vibration events pattern recognition used in fiber fence

设计了一种用于光纤围栏扰动信号模式识别的方法,详细描述了其中的数据完整性确认、特征提取、特征库、分类器等模块;设计了三组实验,对模式识别效果进行了验证。实验结果表明:在光纤围栏系统中引入模式识别模块,能在保证告警率的同时有效降低误告警率。

Patterns recognize method of vibration signal classification used in fiber fence was designed.Component parts of this method as feature extractor,feature database and feature classification was described.Three groups of experiments in proving effect of vibration signal classification was wrote up.The result of experiments show that pattern recognition model in fiber fence effectively reduced the false alarm rate while kept alarm rate.