CAJ | 학술논문

由单颗脉冲星定义的脉冲星时受多种噪声源的影响,其短期和长期稳定度都不够好.为了削弱这些噪声源对单脉冲星时的影响,可以采取合适的算法对多个单脉冲星时进行综合得到综合脉冲星时,从而提高综合脉冲星时的长期稳定度.文中介绍4种综合脉冲星时算法:经典加权算法、小波分析算法、维纳滤波算法和小波域中的维纳滤波算法,将这4种算法分别应用于Arecibo天文台对两颗毫秒脉冲星PSR B1855+09和PSRB1937+21观测得到的计时残差并作出比较.
유단과맥충성정의적맥충성시수다충조성원적영향,기단기화장기은정도도불구호.위료삭약저사조성원대단맥충성시적영향,가이채취합괄적산법대다개단맥충성시진행종합득도종합맥충성시,종이제고종합맥충성시적장기은정도.문중개소4충종합맥충성시산법:경전가권산법、소파분석산법、유납려파산법화소파역중적유납려파산법,장저4충산법분별응용우Arecibo천문태대량과호초맥충성PSR B1855+09화PSRB1937+21관측득도적계시잔차병작출비교.
Pulsars, rapidly rotating neutron stars, have extremely stable rotating periods. The pulsar time denned by a single pulsar is influenced by several noise resources. To weaken these influences, the ensemble analysis method is used to obtain the ensemble pulsar time so that the long-term stability of the ensemble pulsar time can be improved. In this paper, four algorithms — the classical weighted average algorithm, the wavelet analysis algorithm, the Wiener filtration analysis algorithm and the Wiener filtration analysis in wavelet domain — are applied to synthetically make an ensemble pulsar time. The data used are the residuals of the two millisecond pulsars (PSR B1855+09 and PSR B1937+21) observed by Arecibo Observatory. First, the classical weighted average algorithm is developed by Petit, in which only one weight can be chosen within the whole interval of the observation on each single pulsar time, and the criterion for weight is the stability σ_x~2(T) of each single pulsar time. Second, an ensemble pulsar time algorithm is developed based on the wavelet multi-resolution analysis and the wavelet packet analysis, which can be obtained by decomposing the observation residuals of pulsars, extracting the components of different frequency domain and then choosing the weight according to the stability of different component denoted with wavelet variance. Third, the pulsar timing residuals are caused by reference atomic clock and pulsar itself, which are uncorrelated. Considering this uncorrelation and the peculiarity of Wiener filtration, we put forward an ensemble pulsar time algorithm of Wiener filtration. Using this algorithm, the error can be separated from an atomic clock and the pulsar itself in the post-fit pulsar timing residuals. The atomic scale component can be filtered from the pulsar phase variations and the remains can be integrated to the ensemble pulsar time. Weights are chosen according to the mean square root. Forth, the wavelet analysis and the Wiener filtration algorithm are combined and a new ensemble pulsar time algorithm called the Wiener filtration analysis in wavelet domain is presented in this paper, which can remove the influence of noise more effectively. The pulsar timing residuals are decomposed to different components by wavelet. The influences of the atomic clock of different components are removed by Wiener filtration and then the ensemble pulsar time can be obtained by inverting the wavelet transform acting on these remains. The computed result indicates that the latter three algorithms are far better than the first one and the Wiener filtration analysis in wavelet domain could be the best one among all algorithms.