| |
 |
UNIVERSITY OF BUCHAREST
FACULTY OF PHYSICS Guest
2024-11-22 2:04 |
 |
|
|
|
Conference: Bucharest University Faculty of Physics 2021 Meeting
Section: Atomic and Molecular Physics. Astrophysics. Applications
Title:
Gravitational Waves and Noise Classification Using Neural Networks
Authors:
Andrei-Ieronim CONSTANTINESCU(1,2), Ana CARAMETE(1), Laurentiu-Ioan CARAMETE(1)
*
Affiliation:
1) Institute of Space Science, Măgurele, Romania
2) University of Bucharest, Faculty of Physics, Măgurele, Romania
E-mail
aiconstantinescu@spacescience.ro
Keywords:
gravitational waves, noise, neural network
Abstract:
Gravitational waves (GW) have gained a lot of visibility since the first official gravitational wave has been detected in 2015 by the Ligo-Virgo Collaboration. Currently there are only ground-based gravitational waves detectors. LISA (Laser Interferometry Space Antenna) which is planned to be launched in 2034 will be a GW observer in space consisting of three satellites with a range of 2.5 million kilometers between them that are placed in a triangle with different arm lengths. The two main objectives of this mission, among many other, are to detect gravitational waves as well as to emit alerts to other observatories for multi messenger astronomy purposes, meaning we want to observe the event that emits the gravitational waves from the gravitational and electro-magnetic point of view. In order to achieve these main objectives a fast signal analysis method is required on board of the satellites. Neural network (NN) algorithms have been approved for this kind of necessities. In this paper it is presented a recurrent bi-directional long-short term memory neural network that is capable to characterize and classify GW signals and noises. This NN characterizes GW signals and noise based on two spectral characteristics (instant frequency and spectral entropy), and classifies them by the source parameters. The NN is afterwards tested on a month worth set of data consisting mainly noise and just one GW signal in order to observe its performance.
References:
[1] Karsten Danzmann, LISA Laser Interferometer Space Antenna A proposal in response to the ESA call for L3 mission concepts, https://www.elisascience.org/files/publications/LISA_L3_20170120.pdf
[2] A. Caramete, A. I. Constantinescu, L. I. Caramete, T. Popescu, R. A. Balasov, D. Felea, M. V. Rusu, P. Stefanescu, O. M. Tintarean, ‘Characterization of Gravitational Waves Signals Using Neural Networks’, arXiv, 2009.06109, 2020
|
|
|
|