原子时的小波分解算法

An Atomic Time Scale Algorithm Using Wavelet Decomposition

时间尺度的建立就是根据一组原子钟的数据 ,用统计的方法计算出平均的时间尺度。其目的是使综合时间尺度的噪声最小。一般的原子时算法 ,都不能脱离经典加权的局限性 ,只能抑制原子钟的某一种噪声。运用小波分解的方法建立小波分解原子时 ,将原子钟信号在小波域按频率尺度分解 ,然后在不同频率范围加权 ,这种方法不但考虑不同原子钟在同一个频率分量的不同稳定度 ,而且考虑不同频率分量的不同稳定度 。

The weight of atomic clock is assigned traditionally based on its historical performance. The algorithms pay attention to the long term instability of atomic clock and can't respond to the change of clock. In this paper,an algorithm for atomic time scale is proposed based on wavelet decomposition method. For this algorithm,the clock signal is decomposed in various frequency range,the instability is expressed with wavelet variance, and the clock is weighted for various frequency range. This method takes into consideration both short term and long term instabilities of atomic clock, and it is better than other algorithms. This algorithm has been tested with the data from Shaanxi Astronomical Observatory.