LMS算法在分布式光纤测温系统中的应用

Application of LMS Algorithm in the Distributed Optical Fiber Temperature Measurement System

信号采集和数据处理是分布式光纤测温系统的核心,信号中的噪声决定了整个系统的测量精度,空间分辨率,采集速度及最终的请求响应时间。通常系统采集到微弱的温度信息的反斯托克斯信号会被完全淹没在噪声中。因此,得到含有温度信息的有用信号就显得尤为重要。文中针对系统中噪声信号的特征,提出一种变步长的最小均方根自适应滤波算法,对拉曼散射的温度信号进行去噪处理并进行Matlab仿真。仿真结果表明,该算法有效提高了去噪的效果,从而可提升整个系统的测量精度,空间分辨率。

The core of a distributed optical fiber temperature measurement system （DTS） is signal collection and data processing. The measuring accuracy, spatial resolution, collecting speed and final response time are all af- fected by the noise in signal. The anti-Stokes signal contains very weak temperature information and is often com- pletely overwhelmed by noise. Therefore, getting the useful temperature information from the signal is particularly important. A variable step size least-mean-square （LMS） adaptive filtering algorithm is researched for the character- istic of the noise signal in distributed optical fiber temperature measurement system and used to denoise the Raman scattering temperature signal. Matlab simulation shows that this algorithm has good denoising performance and im- proves the system measurement accuracy and spatial resolution.