Logo image
Building numerical relativity surrogate models with neural networks: a thesis in Physics
Thesis   Open access

Building numerical relativity surrogate models with neural networks: a thesis in Physics

Tyson C. George
Master of Science (MS), University of Massachusetts Dartmouth
2023
DOI:
https://doi.org/10.62791/20306

Abstract

With the detection of gravitational waves and use of numerical relativity (NR), we are able to study the properties of powerful astrophysical events. However, due to the complexities of solving Einstein’s equations, other techniques such as surrogate modeling have come into fruition. Surrogate models offer comparable accuracy to that of their NR counterpart, while allowing for waveform evaluations in a fraction of the time. However these too fall short when performing parameter estimations; which can have upwards of millions of waveform evaluations, ramping up total run-time. To mitigate this, we employ the use of GPU-accelerated neural networks, which offer a significant speed-improvement. Using neural networks, we have built a 1D model trained from the hybrid surrogate NRHybSur3dq8 where we target the dominant ℓ = m = 2 mode. Building over the mass-ratio range q ∈ [1, 10], we analyze the overall accuracy in waveform generation and compare total run-times. Further work towards building a 3D model and attaining feature-parity with the surrogate is also explored.
pdf
George T.C. COE MS Thesis 20231.65 MBDownloadView
CC BY-NC-ND V4.0 Open Access

Metrics

7 Record Views

Details

Logo image