Noise Analysis And Application Of Millisecond Pulsar Timing

Pulsars are rapidly rotating neutron stars with extremely stable rotation periods,and based on the long-term observation and analysis arrival time series of pulse,a wealth of information about pulsar basic parameters,interstellar medium and stellar evolution can be obtained.Millisecond pulsars,in particular,have a high rotation period stability,which can be applied to such things as examining gravitational waves,pulsar navigation and providing time standards after years of observation and obtaining a large amount of timing data.However,millisecond pulsar timing data are also affected by the unknown generation mechanism,irregular changes and ubiquitous timing noise,which is very limiting the application of millisecond pulsar with long-term timing data.In order to avoid the influence of timing noise,among many related applications,millisecond pulsars with low timing noise are preferred to be studied,which has directly contributed to a substantial decrease in the number of available pulsars.At present,the research on timing noise analysis and noise reduction methods of millisecond pulsar are not perfect,and many problems need to be solved.This paper mainly studies these problems,including: Detection of timing noise of millisecond pulsar,spectral structure analysis of timing noise of millisecond pulsar and how to avoid the influence of timing noise.This paper also won the support of national natural fund,for example: “noise analysis and processing on pulsar time scale”,“studies in pulsar timing error based on radio observation”,“establishment of pulsar time scale and its application in time keeping theory”,“ research on application of pulsar timescale”,etc.The main work of this paper can be summarized as follows:In the first part,the pulsar observation technology,the principle and method of timing observation are summarized.The pulsar phase transition model and pulse arrival time analysis model are analyzed.Various error sources affecting pulsar timing observation are analyzed,including pulsar rotation instability error,interstellar dispersion delay error,gravitational wave background radiation error,atomic time error and solar system planetary ephemeris error.The second part studies the identification method of millisecond pulsar timing noise.Several measurement models of timing noise are analyzed.Based on the fact that the timing noise of millisecond pulsars is generally small and difficult to be recognized,a Hurst index method is proposed.Hurst index can effectively identify the intensity of millisecond pulsar timing noise.Considering the influence of observation time and residual dispersion on timing noise,their internal relations are analyzed.At the same time,the difference between Hurst index method and other models is given.The effective scheme of timing noise recognition for millisecond pulsar is presented.In the third part,the contribution of noises with different time scales is studied through the decomposition of millisecond pulsar timing residuals,and the prediction of timing residuals is also carried out.According to the irregular fluctuation characteristics of millisecond pulsar timing residuals,a highly adaptive empirical mode decomposition(EMD)method is proposed.Thus,the conversion of random residual fluctuation to simple and orderly is established,which is conducive to the frequency analysis of pulsar data in the later stage and the corresponding power spectrum is obtained.At the same time,EMD method was used to study the difference of residual wave characteristics of different pulsars,and corresponding explanation is given.Furthermore,the method of residual sequence prediction based on EMD method is studied,and the advantages of EMD method in pulsar residual prediction are obtained.In the fourth part,we study a new algorithm to establish ensemble pulsar timescale based on the purpose of noise reduction.By revealing the co-integration relationship among some residual sequences of pulsars,an algorithm based on cointegration theory is proposed to establish the ensemble pulsar time scale.Using pulsar data,the effectiveness of this algorithm in reducing timing noise is verified.This method is applied to the establishment of pulsar time scale,which can solve the problem that pulsar data with relatively strong timing noise is limited in time scale algorithm due to poor stability.