自适应小波包的光纤传感器信号压缩感知

The compressed sensing of optical fiber sensor signal based on adaptive wavelet packet

摘要:

在由分布式光纤传感器构成光纤周界报警系统中 ,针对大规模、高分辨率的光纤振动信号在采样、传输、存储和重 构过程中会受到网络带宽、存储容量等限制问题,提出基于自适应小波包的光纤传感器信号 压缩感知方法。首先,采用小波 包对光纤振动信号进行多层稀疏变换,通过求取小波包系数的数学期望来选取初始置零阈值 ；然后,对光纤振动信号进行重 构并求取重构信号精度,再根据信号的重构精度采用迭代计算方法求取小波包系数的最佳置 零阈值,使光纤振动信号在满足 重构精度的前提下具有最高稀疏度,实现信号的自适应压缩感知；最后,根据光纤振动信号 的特征,采用K-SVD算法训练 得到过完备字典,结合正交匹配跟踪方法完成光纤振动信号的高精度重构。大量实验证明, 与传统压缩算法相比较,新方法的各方面性能均得到较大程度 提高。

Abstract:

In the optical fiber perimeter alarm system composed of distributed op tical fiber sensors,for the problem that large and high-resolution optical fiber vibration signals will be limited by network bandwidth and storage capacity in the process of sampling,transmission,storage and reconstruction,a compressed sensing method of optical fiber sensor signal based on adaptive wavelet packet is proposed.Firstly,the optical fiber vibration signals are mul ti-layer sparsity transformed using wavelet packet,and the initial zero threshold is selected by calculating the mathematical expectati on of the wavelet coefficient; Then,the optical fiber vibration signal is reconstructed and the reconstruction signal precision is obt ained,according to the signal reconstruction precision, the optimal zero threshold value of wavelet coefficient is obtained by using the iterative calculation method,so that the vibration signal has the highest sparsity degree under the premise of satisfying the recon struction accuracy,and the adaptive compression perception of the signal is realized.Finally,the over-complete dictionary was trained by K-SVD algorithm,and the high-precision reconstruction of optical fiber vibration signal was completed by orthogonal mat ching tracking method.A large number of experiments show that the performance of the new method is greatly improved comp ared with the traditional compression algorithm.