利用传统的洛伦兹线型拟合来进行布里渊散射谱特征提取是一种普遍使用的方法, 其测量结果的准确度容易受噪声含量、扫频谱型和拟合算法的影响。不仅如此, 这种方法需要反复迭代求出参数向量的最小二乘解, 实时性较差。针对该情况提出的互相关卷积与高阶质心结合算法能有效提取布里渊频移, 具有极短的测量时间, 但扫频数据服从洛伦兹分布的程度直接影响了计算结果的准确度, 因此提出了二维小波阈值去噪方案, 并运用综合评价指标计算与拐点定位进行最优参量选取, 得到了一组最佳的降噪参量, 如此能有效提高测量结果的精准性。

It is a common method to extract the feature of Brillouin scattering spectrum using the traditional Lorenz fitting, but the measurement accuracy is easily affected by noise, sweep spectrum and fitting algorithm. Moreover, it requires repeated iterations to find the least squares solution of the parameter vector, so the realtime performance is poor. In this case, the algorithm based on crosscorrelation convolution and higherorder centroid calculation can effectively extract Brillouin frequency shift with very short measurement time, but the degree of sweep data obeying the Lorentz distribution directly affects the accuracy of the calculation results. Therefore, in this paper a twodimensional wavelet threshold denoising scheme was presented and the comprehensive evaluation index calculation and the inflection point positioning were used for the optimal parameter selection to get a set of the best noise reduction parameters. And the use of this denoising scheme can effectively improve the accuracy of measurement results.