A machine learns to predict the stability of circumbinary planets

Abstract

Long-period circumbinary planets appear to be as common as those orbiting single stars and have been found to frequently have orbital radii just beyond the critical distance for dynamical stability. Assessing the stability is typically done either through N-body simulations or using the stability criterion first considered by Dvorak and later developed by Holman and Wiegert: a second-order polynomial calibrated to broadly match numerical simulations. However, the polynomial is unable to capture islands of instability introduced by mean motion resonances, causing the accuracy of the criterion to approach that of a random coin-toss when close to the boundary. We show how a deep neural network (DNN) trained on N-body simulations generated with REBOUND is able to significantly improve stability predictions for circumbinary planets on initially coplanar, circular orbits. Specifically, we find that the accuracy of our DNN never drops below 86%, even when tightly surrounding the boundary of instability, and is fast enough to be practical for on-the-fly calls during likelihood evaluations typical of modern Bayesian inference. Our binary classifier DNN is made publicly available at https://github.com/CoolWorlds/orbital-stability

arXiv commentAccepted by MNRAS. Deep neural network and implementation can be found at https://github.com/CoolWorlds/orbital-stability

References in corpus (17)

• The Transiting Exoplanet Survey Satellite
• The PLATO 2.0 Mission
• REBOUND: An open-source multi-purpose N-body code for collisional dynamics
• batman: BAsic Transit Model cAlculatioN in Python
• The NASA Exoplanet Archive: Data and Tools for Exoplanet Research
• Long-Term Stability of Planets in Binary Systems
• IAS15: A fast, adaptive, high-order integrator for gravitational dynamics, accurate to machine precision over a billion orbits
• A Closely-Packed System of Low-Mass, Low-Density Planets Transiting Kepler-11
• Kepler-47: A Transiting Circumbinary Multi-Planet System
• First order resonance overlap and the stability of close two planet systems
• On the Abundance of Circumbinary Planets
• The Next Generation Transit Survey (NGTS)
• A Machine Learns to Predict the Stability of Tightly Packed Planetary Systems
• Uncovering Circumbinary Planetary Architectural Properties from Selection Biases
• Observational biases for transiting planets
• How bright are the gaps in circumbinary disk systems?
• Transit Clairvoyance: Enhancing TESS follow-up using artificial neural networks

Cited by in corpus (24)

• Predicting the long-term stability of compact multiplanet systems
• Toward an AI Physicist for Unsupervised Learning
• TOI-1338: TESS' First Transiting Circumbinary Planet
• Stability Limits of Circumbinary Planets: Is There a Pile-up in the Kepler CBPs?
• Contemporary machine learning: a guide for practitioners in the physical sciences
• Orbital Stability of Circumstellar Planets in Binary Systems
• Kepler-1661 b: A Neptune-sized Kepler Transiting Circumbinary Planet around a Grazing Eclipsing Binary
• Fundamental limits from chaos on instability time predictions in compact planetary systems
• Using Deep Neural Networks to compute the mass of forming planets
• The CORALIE survey for southern extrasolar planets XIX. Brown dwarfs and stellar companions unveiled by radial velocity and astrometry
• Running The Gauntlet -- Survival of Small Circumbinary Planets Migrating Through Destabilising Resonances
• Long-term stability of planets in and around binary stars
• Sculpting the circumbinary planet size distribution through resonant interactions with companion planets
• Machine Learning Classification of Kuiper Belt Populations
• A Pluto--Charon Sonata: Dynamical Limits on the Masses of the Small Satellites
• How the presence of a giant planet affects the outcome of terrestrial planet formation simulations
• Machine learning prediction for mean motion resonance behaviour -- The planar case
• Analyzing the Habitable Zones of Circumbinary Planets Using Machine Learning
• Machine learning approach for mapping the stable orbits around planets
• Dynamical Fates of S-Type Planetary Systems in Embedded Cluster Environments
• Orbital Evolution of a Circumbinary Planet in a Gaseous Disk
• Intelligence, physics and information -- the tradeoff between accuracy and simplicity in machine learning
• Employing machine learning for theory validation and identification of experimental conditions in laser-plasma physics
• The Number of Transits Per Epoch for Transiting Misaligned Circumbinary Planets