论文部分内容阅读
针对光纤布拉格光栅(FBG)传感信号难以去除噪声干扰及信号丢失问题,采用压缩感知(CS)对传感信号进行处理。CS重构算法多是以稀疏度已知为先验条件,提出稀疏度确定方法,结合二次正交匹配追踪(TOMP)算法和广义正交匹配追踪(GOMP)算法提出广义二次正交匹配追踪(GtOMP)算法,确定每次迭代选择原子个数及迭代次数。首先计算相关系数,归一化后按降序排列,并结合饱和值的方法确定稀疏度,利用平稳度找出每次迭代所选择的原子个数,最后利用本文方法对FBG信号进行重构。实验仿真表明,与同类的TOMP算法相比,本文算法不仅运行时间大大减少,而且降低了6~20%的重构误差;与其他不同类算法相比,本文算法重构信号的信噪比(SNR)提高27dB以上。
It is difficult to remove the noise interference and signal loss problem of fiber Bragg grating (FBG) sensing signal, and compressed sensing (CS) is used to process the sensing signal. Most of the CS reconstruction algorithms are based on the known sparsity as a priori condition, and the sparseness determination method is proposed. Combined with the quadratic orthogonal matching pursuit (TOMP) algorithm and the generalized orthogonal matching pursuit (GOMP) algorithm, the generalized quadratic orthogonal matching Tracking (GtOMP) algorithm to determine the number of atoms and the number of iterations for each iteration. Firstly, the correlation coefficient is calculated, then normalized and then arranged in descending order, combined with the saturation method to determine the sparseness, the smoothness is used to find the number of atoms selected for each iteration, and finally the FBG signal is reconstructed by this method. The experimental results show that compared with the same TOMP algorithm, the proposed algorithm not only reduces the running time but also reduces the reconstruction error by 6-20%. Compared with other algorithms, the signal to noise ratio SNR) increased by 27dB or more.