摘要
相位敏感光时域反射计（φ-OTDR）全分布式光纤振动传感系统,在对车辆信号识别过程中所采用的幅度差分法存在着识别率低的问题。为了克服这一问题,将车辆识别方法改进为基于BP神经网络的识别算法。该算法首先利用φ-OTDR系统采集瑞利后向散射光信号,从中提取出车辆振动信号,然后通过主成分分析法（PCA）对振动信号进行降维和特征提取,最后利用BP神经网络模型对高速公路上的车辆进行识别。实验结果表明：利用BP神经网络算法对车辆进行识别,识别率可以达到98.33%。
For the phase sensitive optical time domain reflectometer （φ-OTDR） distributed optical fiber vibration sensing system, there exists the problem of low recognition rate in the amplitude difference method used in vehicle signal recognition. In order to overcome this problem, the vehicle recognition method based on the BP neural network recognition algorithm is proposed. Firstly, the Rayleigh backscattering optical signal is collected by the φ- OTDR system and the vehicle vibration signal is extracted. Then the principal component analysis （PCA） is used to reduce the dimension of the vibration signal and extract the feature. Finally, using the BP neural network to identify vehicles. The experimental results show that the recognition rate of the vehicles can reach 98.33% by using the BP neural network algorithm.
作者
熊显名
崔向良
XIONG Xianming;CUI Xiangliang(Guilin University of Electronic Technology, Guilin Guangxi 541004, China)
出处
《激光杂志》
北大核心
2018年第6期70-73,共4页
Laser Journal
基金
国家自然科学基金(No.61565004)
桂林市科学研究与技术开发课题(No.20150133-3)
关键词
全分布式光纤振动传感系统
Φ-OTDR
BP神经网络
幅度差分法
fully distributed optical fiber vibration sensor system
Φ- OTDR
BP neural network
amplitude difference method
