利用平稳小波变换处理FBG传感信号  

Treatment of FBG sensor signals by stationary wavelet transform

作　　者：何玉钧[1] 王梓蒴[1] 尹成群[1] 李永倩[1] 

机构地区：[1]华北电力大学电子与通信工程系,河北保定071000

出　　处：《红外与激光工程》2011年第11期2295-2300,共6页Infrared and Laser Engineering

基　　金：国家863计划(2006AA092141)

摘　　要：FBG传感系统中,噪声的存在严重影响了解调系统中心波长检测的精确度。小波变换具有强大的时频分析能力,利用平稳小波变换对FBG反射谱信号进行了去噪处理,并与普通离散小波变换的去噪效果进行了对比。结果表明:对于实际FBG反射谱,该方法信噪比要比普通离散小波变换高出3 dB,而均方差仅为普通离散小波变换的1/4。而且该方法保证了消噪后FBG反射谱信号峰值即对应中心波长,克服了普通离散小波变换因下采样操作而造成的非线性相位和非平移不变的缺点。Signals in fiber Bragg grating （FBG） sensing demodulation system always have noises, which deteriorate the detection accurate of demodulation result. Wavelet transform was chosen to process the FBG reflection spectrum signals because of its powerful ability in time-frequency analysis. Stationary wavelet transform was applied to de-noising FBG reflection spectrum signals, and compared with the de- noising effect by using ordinary discrete wavelet transform. The result shows that the actual FBG reflection spectrum＇s signal-to-noise ratios（SNR） by using this method is 3 dB l＇figher than the SNR using ordinary discrete wavelet transform. Simultaneously, its mean squared error （MSE） is only a quarter of the MSE using ordinary discrete wavelet transform. And this method can effectively ensure the Bragg wavelength of de-noising FBG reflection spectrum at the spectrum peak, and overcome the shortcoming of non-linear phase and non-shift-invariant which is caused by sub-sample process in ordinary discrete wavelet transform.

关 键 词：平稳小波 小波变换 FBG传感 去噪 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]