Stability time-scale prediction for main-belt asteroids using neural networks

Abstract

Many asteroids move in the belt between the orbits of Mars and Jupiter under the gravitational attraction of the Sun and planets in the Solar system. If one of these asteroids does not leave the belt during a period, it is considered to be temporarily stable on that time-scale. This paper aims to study the time-scales on which asteroids could stay in the main belt. A simplified situation is studied in which the initial orbital inclinations and the longitudes of the ascending nodes of the asteroids are set to zero. Numerical integration is used to study the temporal stability of the main-belt asteroids. In the integration, the distribution of the instability time for randomly generated particles can be fitted with a function. Thus a reasonable method is presented to choose an integration time-scale based on the percentage of the already unstable particles in relation to all unstable particles. A total of 151 000 particles are generated and then integrated for 8.09 × 106 yr. The integration data are used to train the probabilistic neural networks to predict the stability of particles. A temporal stability map in the a-e plane is obtained from the prediction results of the neural networks.

Machine Learning applied to asteroid dynamics

“…Neural networks can be used to predict the orbital stability of asteroids in the main belt. Liu et al (2021) used ANN to fit data obtained from the integration of 151,000 particles. A temporal stability map in the (a, e) plane was obtained as a result.…”

“…The community working in the area is less established, with just four groups with published papers so far. The use of more advanced techniques is still in its infancy, with just two papers using neural networks (Liu et al (2021) and Carruba et al (2021b)). Finally, fewer papers produced scientific outcomes that can be categorized as belonging to generation or discovery, and none reached the stage of insight.…”

Fast, Simple, and Accurate Time Series Analysis with Large Language Models: An Example of Mean-motion Resonances Identification

“…Machine-learning (ML) techniques are now actively used in astronomy and celestial mechanics. Using ML, it is possible to identify asteroids trapped in mean-motion resonances (MMRs; Smirnov & Markov 2017), predict orbital stability (Liu et al 2021) or long-term stability of planetary systems (Lam & Kipping 2018;Tamayo et al 2020), find new asteroid families members (Carruba et al 2020), classify Kuiper Belt objects (Smullen & Volk 2020), detect exoplanets (Malik et al 2022), and analyze images (Carruba et al 2021b). Primarily, these study utilize what is often called classical ML, which uses statistical algorithms or models to analyze data.…”