在1916年，也就是广义相对论(GR)场方程最终公式化的一年之后，阿尔伯特·爱因斯坦预测了引力波(GW)的存在，GW是由大质量物体以极快的加速度运动而引起的时空波动。1974年，赫尔斯和泰勒发现双中子星系统PSR B1913+16的轨道由于辐射引力波而收缩，间接证明了GW的存在。值得注意的是，2015年9月14日，LIGO探测器首次直接探测到GW，并首次观测到双黑洞(BBH)合并，开启了研究宇宙的新窗口。因此，研究这些GW事件不仅可以检验引力和天体物理学的理论，如GR，而且还可以向我们展示在双星系统中致密天体的各种组合的全貌，这将带来对宇宙的新认识。
     在本文中，我们将贝叶斯分析和马尔可夫链蒙特卡洛(MCMC)模拟应用于15个银河系双中子星(DNS)系统，并得到其子星质量应该主要落在1.165$-$1.590$M_{\odot}$，以及其总质量、质量比和啁啾质量（$\mathcal{M}$）的主要分布范围分别是2.535$-$2.867$M_{\odot}$, 0.741$-$0.995 和 1.115$-$1.237$M_{\odot}$。我们的结果与GW170817中DNS的性质一致，其子星质量、总质量、质量比和啁啾质量的置信区间（90\%）分别为 1.16$-$1.60 $M_\odot$, $2.73_{-0.01}^{+0.04} M_\odot$, 0.73$-$1.00 和 $1.186_{-0.001}^{+0.001} M_\odot$。上述相似性是一个重要的指标，它证实了GW170817的来源是一个来自星系NGC 4993的DNS系统。正在运行的GW O3探测可以检验我们的结果。
      其次，我们在两个合成模型(模型 Galaxy 和 模型 LIGO)中测试了LIGO-Virgo使用的啁啾质量($\mathcal{M}$)剪切规则的有效性，其中黑洞和中子星的质量分别来自我们星系的电磁波谱观测和LIGO-Virgo的引力波观测(O1和O2)。模拟结果表明，由于LIGO-Virgo所推断的黑洞通常比我们星系的黑洞要大，因此上述剪切规则不适用于模型LIGO，而NS-BH事件的$\mathcal{M}$将分布在$2.1M_{\odot}<\mathcal{M}<7.3M_{\odot}$的范围内，它与BH-BH事件有部分重叠。因此，我们建议在新的LIGO-Virgo搜寻轮(例如O3)中，应仔细寻找在 $2.1M_{\odot}<\mathcal{M}<7.3M_{\odot}$范围内潜在的NS-BH候选体。并且，我们利用上述合成模型(即模型 LIGO)，采用蒙特卡罗（MC）方法对标准Novikov-Thorne吸积盘模型进一步模拟了NS-BH合并的GW频率($f_{\mathrm{GW}}$)的分布。我们的结果预测，在史瓦西黑洞的情况下，当系统刚好在合并前，$f_{\mathrm{GW}}$的置信度为90\%的区间是$165_{-64}^{+475}$ Hz，这个GW频率预计将在合并阶段增加数倍，它位于LIGO-Virgo的频带内，也就是大约15 Hz 到几千 Hz。我们的结果为LIGO-Virgo正在进行的O3观测寻找NS-BH合并提供了一个重要的参考。
       然后，我们利用经典的高斯混合模型(GMMs)对GWTC-1目录中的黑洞质量进行分析，以确定LB-1和GW190521中报道的大质量黑洞是否与GWTC-1目录中的 LIGO-Virgo 黑洞(O1和O2)的种群一致。结果表明，LB-1和GW190521中已报道的大质量黑洞与我们星系和GWTC-1里的黑洞并不一致，它们可能属于与正常黑洞不同起源的新一类黑洞。最后，我们对全文进行了总结，并对今后的工作进行了展望。

In 1916, the year after the final formulation of the field equations of general relativity (GR), Albert Einstein predicted the existence of gravitational waves (GW) that are ripples in space-time caused by massive objects moving with extreme accelerations. The discovery of the binary pulsar system PSR B1913+16 by Hulse and Taylor in 1974 and subsequent observations of its energy loss indirectly demonstrated the existence of GW. Significantly, on 14 September 2015, the first direct detection of GW and the first observation of a binary black hole (BBH) merger were made by LIGO detector, in which it has opened a new window on the universe. Therefore, investigating these GW events not only can test the theories on gravitation and astrophysics, such as GR, but also can show us the whole pictures of various combinations of compact objects in binary systems that should bring out the new insights into the Universe.
  In this thesis, we apply a Bayesian analysis and Markov Chain Monte Carlo (MCMC) simulation to 15 Galactic Double Neutron Star (DNS) systems, and obtain that the component masses of DNS systems should mainly fall in the range of 1.165$-$1.590$M_{\odot}$, and the predominant ranges for the total mass, mass ratio and chirp mass lie in 2.535$-$2.867$M_{\odot}$, 0.741$-$0.995 and 1.115$-$1.237$M_{\odot}$, respectively. Our results are in agreement with the properties of DNS in GW170817, whose 90\% credible intervals for the component masses, total masses, mass ratio and chirp masses are 1.16$-$1.60 $M_\odot$, $2.73_{-0.01}^{+0.04} M_\odot$, 0.73$-$1.00 and $1.186_{-0.001}^{+0.001} M_\odot$, respectively. The above similarity is an important indicator that reveals the source of GW170817 to be a DNS system from the galaxy NGC 4993, and our results could be tested by the forthcoming GW hunting O3.
  We test the validity of the cut rule of chirp mass ($\mathcal{M}$) used by LIGO-Virgo in two synthetic models, i.e. Model Galaxy and Model LIGO, in which the masses of BHs and NSs are observed by the electromagnetic spectrum observations in our Galaxy and inferred by LIGO-Virgo detection (O1 and O2), respectively. The simulation shows that it is unsuitable for Model LIGO due to the BHs inferred by LIGO-Virgo are usually bigger than those in our Galaxy, and $\mathcal{M}$ of NS-BH events would distribute in the range of $2.1M_{\odot}<\mathcal{M}<7.3M_{\odot}$, which partially overlaps with those of BH-BH events. Therefore, we suggest that the new searching round of LIGO-Virgo (e.g. O3) should carefully seek out the underlying NS-BH candidates in the range of $2.1M_{\odot}<\mathcal{M}<7.3M_{\odot}$. Moreover, we further simulate the GW frequency ($f_{\mathrm{GW}}$) distribution of NS-BH mergers by above synthetic model ( i.e. Model LIGO), with the standard Novikov-Thorne accretion disk model through the Monte Carlo method. Our results predict that the median with 90\% credible intervals for which is $165_{-64}^{+475}$ Hz in the case of Schwarzschild BH when the system just before merger, and this GW frequency is expected to increase several times in the merger stage, which is lying in the frequency band of LIGO-Virgo, i.e., about 15 Hz to a few kHz. Our results provide an important reference for hunting the NS-BH mergers by the on-going O3 run of LIGO-Virgo.
  We analyze the BH masses in GWTC-1 catalog with the classical Gaussian mixture models (GMMs) to determine whether the declaimed massive BH in LB-1 or GW190521 is consistent with the same population as the LIGO-Virgo BHs (O1 and O2) in GWTC-1 catalog or not. Our results reveal that the reported massive BHs in LB-1 and GW190521 are inconsistent with the same population as those of our Galaxy and the GWTC-1, and they may belong to the new class of BHs with the other origins different from the normal BHs. In the end, we summarize the paper and put forward the prospects for our future work.