CAJ | 학술논문

本文主要提供一种基于光纤光栅周界报警系统的模式识别方法，其中方法包括以下步骤：当信号振动幅度大于某一阈值时，认为信号发生突变，记录突变信号后一段时间的振动数据，然后求取振动数据极值，并对其做阻力震荡衰减趋势拟合，得到拟合特征参数斜率和相关系数与设定阈值进行比较，从而达到对单次激励行为的识别。本文提供的模式识别方法识别周界动物碰触敏度高，抗环境能力强，系统报警性能好，误识别低。
본문주요제공일충기우광섬광책주계보경계통적모식식별방법，기중방법포괄이하보취：당신호진동폭도대우모일역치시，인위신호발생돌변，기록돌변신호후일단시간적진동수거，연후구취진동수거겁치，병대기주조력진탕쇠감추세의합，득도의합특정삼수사솔화상관계수여설정역치진행비교，종이체도대단차격려행위적식별。본문제공적모식식별방법식별주계동물팽촉민도고，항배경능력강，계통보경성능호，오식별저。
This paper provides a kind of pattern recognition method based on fiber Bragg grating perimeter alarm system.Method includes the following steps: when the signal is greater than a certain threshold,the signal is considered to mutate, record mutation signal after a period of vibration data,then vibration data extreme value is calculated,and its resistance to shock attenuation trend fitting.Characteristic parameters by fitting the slope and the correlation coefficient comparing with setting threshold,so as to achieve the recognition of a single motivation behavior.This article provides a pattern recognition method to identify perimeter animals touch sensitive degree is high,strong ability to resist the environment,the system alarm performance is good,low error identification.