Conditional autoencoder for generating binary neutron star waveforms with tidal and precession effects
条件自编码器生成含潮汐和进动效应的双中子星波形
Gravitational waves from binary neutron star mergers provide insights into dense matter physics and strong-field gravity, but waveform modeling remains computationally challenging. We develop a deep generative model for gravitational waveforms from binary neutron star mergers, covering the late inspiral, merger, and ringdown, incorporating precession and tidal effects. Using the conditional autoencoder, our model efficiently generates waveforms with high precision across a broad parameter space, including component masses <inline-formula><mml:math><mml:mo>(</mml:mo><mml:msub><mml:mi>m</mml:mi><mml:mn>1</mml:mn></mml:msub><mml:mo>,</mml:mo><mml:msub><mml:mi>m</mml:mi><mml:mn>2</mml:mn></mml:msub><mml:mo>)</mml:mo></mml:math></inline-formula>, spin components <inline-formula><mml:math><mml:mo>(</mml:mo><mml:msub><mml:mi>S</mml:mi><mml:mrow><mml:mn>1</mml:mn><mml:mi>x</mml:mi></mml:mrow></mml:msub><mml:mo>,</mml:mo><mml:msub><mml:mi>S</mml:mi><mml:mrow><mml:mn>1</mml:mn><mml:mi>y</mml:mi></mml:mrow></mml:msub><mml:mo>,</mml:mo><mml:msub><mml:mi>S</mml:mi><mml:mrow><mml:mn>1</mml:mn><mml:mi>z</mml:mi></mml:mrow></mml:msub><mml:mo>,</mml:mo><mml:msub><mml:mi>S</mml:mi><mml:mrow><mml:mn>2</mml:mn><mml:mi>x</mml:mi></mml:mrow></mml:msub><mml:mo>,</mml:mo><mml:msub><mml:mi>S</mml:mi><mml:mrow><mml:mn>2</mml:mn><mml:mi>y</mml:mi></mml:mrow></mml:msub><mml:mo>,</mml:mo><mml:msub><mml:mi>S</mml:mi><mml:mrow><mml:mn>2</mml:mn><mml:mi>z</mml:mi></mml:mrow></mml:msub><mml:mo>)</mml:mo></mml:math></inline-formula>, and tidal deformability <inline-formula><mml:math><mml:mo>(</mml:mo><mml:msub><mml:mi>Λ</mml:mi><mml:mn>1</mml:mn></mml:msub><mml:mo>,</mml:mo><mml:msub><mml:mi>Λ</mml:mi><mml:mn>2</mml:mn></mml:msub><mml:mo>)</mml:mo></mml:math></inline-formula>. Trained on <inline-formula><mml:math><mml:mn>1</mml:mn><mml:mo>×</mml:mo><mml:msup><mml:mn>10</mml:mn><mml:mn>6</mml:mn></mml:msup></mml:math></inline-formula> waveforms from the IMRPhenomXP_NRTidalv2 waveform model, our model achieves a mean mismatch of <inline-formula><mml:math><mml:mn>2.13</mml:mn><mml:mo>×</mml:mo><mml:msup><mml:mn>10</mml:mn><mml:mrow><mml:mo>-</mml:mo><mml:mn>3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>. The model accelerates waveform generation. For a single sample, it requires 0.12 s (s), compared to 0.66 s for IMRPhenomXP_NRTidalv2 making it approximately 5 times faster. When generating 1000 waveforms, the network completes the task in 0.75 s, while IMRPhenomXP_NRTidalv2 requires 7.12 s, making it approximately 10 times faster. This speed advantage enables rapid parameter estimation and real-time gravitational wave searches. With higher precision, it will support low-latency detection and broader applications in multimessenger astrophysics.
展开 ▾首次使用条件自编码器同时处理双中子星波形的自旋进动和潮汐变形,平均不匹配度仅2.13×10⁻³,并揭示了潜在空间的物理可解释性,实现了高速高精度波形生成。
自LIGO首次探测到双黑洞并合Abbott+ 2016和双中子星并合GW170817Abbott+ 2017以来,对波形建模的速度和精度要求日益提升。传统模型如IMRPhenomXP_NRTidalv2Colleoni+ 2025虽然融合了进动与潮汐效应,但计算耗时;深度学习通过George+ 2018、Dax+ 2025、Whittaker+ 2022、Chua+ 2021等工作在加速波形生成上取得了显著进展,然而多未同时包含进动和潮汐。本文引入条件自编码器,以该解析模型波形为训练数据,实现了兼顾自旋进动和潮汐变形的快速高精度波形生成。未来可借助TensorRTChaturvedi+ 2022和ONNX RuntimeAlizadeh & Castor 2024进一步优化推理速度,并探索扩散模型Mo & Tian 2024等架构向更复杂物理系统扩展。
预印本 2025-03-25 · 刊出 2025-10-07 · 收录 2026-10-13