基于快速加权l_1范数最小化的FBG信号去噪方法

FBG Signal De-noising Method Based on Weighted l1-norm Minimization

光纤布拉格光栅(FBG)信号去噪是实际FBG传感系统的核心问题。现有的方法可以达到较高的信噪比,但不能保证较低的峰值位置漂移,这是FBG传感系统的关键。文章提出了一种基于快速加权l_1范数最小化的FBG信号去噪方法,该方法基于加速近端梯度算法,具有低DPP的特点,去噪后峰值位置几乎不变。仿真结果表明,该算法具有快速高效的去噪效果,对参数选择具有较强的鲁棒性,而且文章算法只涉及基本运算操作,易于应用在实际处理系统中。

FBG signal de-noising is a central problem for constructing apractical FBG sensing system.Existing methods can achieve high SNR while they do not guarantee to keep low displacement of peak position,which is a special requirement of FBG sensing system.In this paper,we propose a novel weighted l1-norm minimization algorithm for FBG sensing signal de-noising.A special advantage of our algorithm is that it has low DPP,which means that the peak position is nearly invariant before and after de-noising,thus it doesn＇t need nonlinear fitting sequentially.Simulation results indicate that the proposed algorithm is fast and efficient in de-noising,and also,it is robust to parameter selection.Besides,this algorithm can be easily implemented in real-time processing systems,since it only needs elementary operations.