卡尔曼滤波在核脉冲全谱数据采集系统中的应用

The application of Kalman filter in the nuclear pulse full spectrum data acquisition system

针对核脉冲全谱数据采集技术中采用的深度加权滤波法在数据量少时平滑效果差、滞后大的缺点,设计了卡尔曼滤波算法替代传统滤波算法进行解谱的方案。滤波算法首先根据信号与噪声的状态空间模型,建立状态方程和测量方程,然后根据广义卡尔曼滤波对测量方程进行更新,最后根据现场的标准刻度井测量数据确定测量矩阵,对测量数据进行滤波。实测结果表明,该滤波算法相对于传统滤波算法,能够消除谱图的统计涨落,提高光滑度,遏制滞后,并保证滤波后曲线不失真。

The weighted depth filtering is used in the nuclear pulse full spectrum data acquisition system. However, this filtering algorithm can cause unsmooth waveform and large lag in small-data phenomenon. In view of this, the Kalman filtering is introduced into the system mentioned above as a replacement algorithm. We build the state equation and measurement equation according to the state-space model of both signal and noise, and then renew the measurement equation with extended Kalman filter. Finally, the measurement matrix is determined according to the logging instrument in the actual application and filter the data observed in the actual measurement. The actual measurement results have shown that the filter can eliminate the statistical fluctuation spectra, improve the smoothness and curb lag on the premise that ensure the curve is not distortion after filtering.