[P-PS08] Lunar Science and Exploration

[PPS08-P10] Development of explainable AI-based prediction model of solar energetic particle events for Moon and Martian exploration

*Chihiro Mitsuda^1,2, Yuta Kato^1,2, Aiko Nagamatsu³, Kanya Kusano², Yoshizumi Miyoshi² (1.Fujitsu Limited, 2.Institute for Space–Earth Environmental Research, Nagoya University, 3.Space Exploration Innovation Hub Center, Japan Aerospace Exploration Agency)

Keywords:Solar Energetic Particles, Explainable AI, Moon and Martian exploration, Space Weather

Fujitsu and the Tokai National Higher Education and Research System jointly research on enhancement of space weather forecasting since February 24, 2023. Solar energetic particles (SEPs) are a kind of cosmic radiation that is originated from solar flares (SFs) and/or Corolan Mass Ejections (CMEs). They are known to affect humans and space systems in lunar and near-Earth space. To further enhance safety for future lunar and Martian exploration activities, we have started further joint research project with the Japan Aerospace Exploration Agency (JAXA) on SEP events predictioin from February 1, 2025.

Our initial approach is to analyzing radiation measurement data from JAXA's lunar surface and lunar orbit missions related to SEP events to identify in-situ particle events. Then we try to apply Fujitsu's explainable AI, "Fujitsu Kozuchi XAI WideLearning" to these events. This model could show the hypotheses which these in-situ particle events are observed.

Our goal is to build a classification prediction model that provides high-accuracy predictions along with their hypothesis for lunar surface, lunar orbit, and deep-space exploration missions, which will require more sophisticated operational decision-making. This presentation reports on the progress of this project.