针对传统的谱线平滑方法在滤波器参数选择不当或平滑次数过多时引起谱线畸变的缺点,在最小二乘法的基础上,利用自适应滤波器的原理,采用快速卡尔曼实现递归最小二乘法,并通过前向、后向预测器更新滤波器系数,实现能谱数据的平滑去噪处理。通过实验将该方法与多项式最小二乘法、最小均方算法进行定性分析与比较,结果表明该方法能较好地降低能谱中的噪声,并能保持能谱的特征。 Aiming at the shortcomings of the conventional spectral line smoothing method when the parameters of the filter are improperly selected or the number of smoothing times is too large, based on the least squares method and the principle of adaptive filter, fast Kalman is used to realize the recursive least squares , And through the forward and backward predictors to update the filter coefficients to achieve smooth spectral data denoising. The method is compared with the polynomial least squares and least mean squares algorithm through experiments. The results show that this method can reduce the noise in the energy spectrum and maintain the characteristics of the energy spectrum.