اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدDetecting the early inspiral of a gravitational-wave signal with convolutional neural networks
76
0
0.0
(
0
)
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
We introduce a novel methodology for the operation of an early %warning alert system for gravitational waves. It is based on short convolutional neural networks. We focus on compact binary coalescences, for light, intermediate and heavy binary-neutron-star systems. The signals are 1-dimensional time series $-$ the whitened time-strain $-$ injected in Gaussian noise built from the power-spectral density of the LIGO detectors at design sensitivity. We build short 1-dimensional convolutional neural networks to detect these types of events by training them on part of the early inspiral. We show that such networks are able to retrieve these signals from a small portion of the waveform.
قيم البحث
اقرأ أيضاً
GW170817 has led to the first example of multi-messenger astronomy with observations from gravitational wave interferometers and electromagnetic telescopes combined to characterise the source. However, detections of the early inspiral phase by the gr
avitational wave detectors would allow the observation of the earlier stages of the merger in the electromagnetic band, improving multi-messenger astronomy and giving access to new information. In this paper, we introduce a new machine-learning-based approach to produce early-warning alerts for an inspiraling binary neutron star system, based only on the early inspiral part of the signal. We give a proof of concept to show the possibility to use a combination of small convolutional neural networks trained on the whitened detector strain in the time domain to detect and classify early inspirals. Each of those is targeting a specific range of chirp masses dividing the binary neutron star category into three sub-classes: light, intermediate and heavy. In this work, we focus on one LIGO detector at design sensitivity and generate noise from the design power spectral density. We show that within this setup it is possible to produce an early alert up to 100 seconds before the merger for the best-case scenario. We also present some future upgrades that will enhance the detection capabilities of our convolutional neural networks. Finally, we also show that the current number of detections for a realistic binary neutron star population is comparable to that of matched filtering and that there is a high probability to detect GW170817- and GW190425-like events at design sensitivity.
Observations of gravitational waves from compact binary mergers have enabled unique tests of general relativity in the dynamical and non-linear regimes. One of the most important such tests are constraints on the post-Newtonian (PN) corrections to th
e phase of the gravitational wave signal. The values of these PN coefficients can be calculated within standard general relativity, and these values are different in many alternate theories of gravity. It is clearly of great interest to constrain these deviations based on gravitational wave observations. In the majority of such tests which have been carried out, and which yield by far the most stringent constraints, it is common to vary these PN coefficients individually. While this might in principle be useful for detecting certain deviations from standard general relativity, it is a serious limitation. For example, we would expect alternate theories of gravity to generically have additional parameters. The corrections to the PN coefficients would be expected to depend on these additional non-GR parameters whence, we expect that the various PN coefficients to be highly correlated. We present an alternate analysis here using data from the binary neutron star coalescence GW170817. Our analysis uses an appropriate linear combination of non-GR parameters that represent absolute deviations from the corresponding post-Newtonian inspiral coefficients in the TaylorF2 approximant phase. These combinations represent uncorrelated non-GR parameters which correspond to principal directions of their covariance matrix in the parameter subspace. Our results illustrate good agreement with GR. In particular, the integral non-GR phase is $Psi_{mbox{non-GR}} = (0.447pm253)times10^{-1}$ and the deviation from GR percentile is $p^{mbox{Dev-GR}}_{n}=25.85%$.
Bayesian model selection provides a powerful and mathematically transparent framework to tackle hypothesis testing, such as detection tests of gravitational waves emitted during the coalescence of binary systems using ground-based laser interferomete
rs. Although its implementation is computationally intensive, we have developed an efficient probabilistic algorithm based on a technique known as nested sampling that makes Bayesian model selection applicable to follow-up studies of candidate signals produced by on-going searches of inspiralling compact binaries. We discuss the performance of this approach, in terms of false alarm rate and detection probability of restricted second post-Newtonian inspiral waveforms from non-spinning compact objects in binary systems. The results confirm that this approach is a viable tool for detection tests in current searches for gravitational wave signals.
We seek to achieve the Holy Grail of Bayesian inference for gravitational-wave astronomy: using deep-learning techniques to instantly produce the posterior $p(theta|D)$ for the source parameters $theta$, given the detector data $D$. To do so, we trai
n a deep neural network to take as input a signal + noise data set (drawn from the astrophysical source-parameter prior and the sampling distribution of detector noise), and to output a parametrized approximation of the corresponding posterior. We rely on a compact representation of the data based on reduced-order modeling, which we generate efficiently using a separate neural-network waveform interpolant [A. J. K. Chua, C. R. Galley & M. Vallisneri, Phys. Rev. Lett. 122, 211101 (2019)]. Our scheme has broad relevance to gravitational-wave applications such as low-latency parameter estimation and characterizing the science returns of future experiments. Source code and trained networks are available online at https://github.com/vallis/truebayes.
The nonlinear aspect of gravitational wave generation that produces power at harmonics of the orbital frequency, above the fundamental quadrupole frequency, is examined to see what information about the source is contained in these higher harmonics.
We use an order (4/2) post-Newtonian expansion of the gravitational wave waveform of a binary system to model the signal seen in a spaceborne gravitational wave detector such as the proposed LISA detector. Covariance studies are then performed to determine the ultimate accuracy to be expected when the parameters of the source are fit to the received signal. We find three areas where the higher harmonics contribute crucial information that breaks degeneracies in the model and allows otherwise badly-correlated parameters to be separated and determined. First, we find that the position of a coalescing massive black hole binary in an ecliptic plane detector, such as OMEGA, is well-determined with the help of these harmonics. Second, we find that the individual masses of the stars in a chirping neutron star binary can be separated because of the mass dependence of the harmonic contributions to the wave. Finally, we note that supermassive black hole binaries, whose frequencies are too low to be seen in the detector sensitivity window for long, may still have their masses, distances, and positions determined since the information content of the higher harmonics compensates for the information lost when the orbit-induced modulation of the signal does not last long enough to be apparent in the data.
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
