カミヤ　トシオ   Toshio KAMIYA   神谷　俊夫    所属   理工学部 総合理工学科    職種   教授
言語種別	英語
発行・発表の年月	2026/10
形態種別	その他の著書・論文
査読	査読あり
標題	A Study on Attitude Control of the LiteBIRD Satellite with Spin and Precession Motion
執筆形態	共著
掲載誌名	Proceedings of the International Astronautical Congress, IAC
掲載区分	国外
担当区分	筆頭著者
著者・共著者	Toshio Kamiya, Hiroki Nakagawa
概要	LiteBIRD is an international space mission designed to search for evidence of cosmic inflation in the early Universe through high-sensitivity measurements of cosmic microwave background (CMB) polarization over the entire sky. The spacecraft will operate around the Sun–Earth Lagrangian point L2 and perform an all-sky survey using a spinning and precessing attitude motion. Because the attitude requirements are still under study, this paper considers representative candidate conditions consisting of a short-term spin period of several minutes, an intermediate-term precession period of several tens of minutes with a prescribed precession angle, and a long-term scan of the precession-axis direction. These widely separated time scales require accurate and computationally efficient long-duration attitude simulation. An attitude-dynamics formulation based on a Lie group variational integrator (LGVI) on SO(3) is developed for a spacecraft equipped with reaction wheels (RWs), including an angular-impulse representation for external thruster inputs. Compared with a conventional fourth-order Runge–Kutta formulation, the LGVI preserves angular momentum, rotational energy, and the SO(3) structure with high accuracy and reduces the computational time by approximately a factor of 20 at a comparable angular-momentum-error level. Based on this formulation, two bias-momentum attitude-control approaches are investigated. In the first, appropriate initial angular velocity and RW bias momentum generate the required spin–precession motion primarily through free rotation, while feed-forward RCS angular impulses control the precession-axis direction. Numerical results show that appropriate RCS firing-phase selection suppresses undesirable changes in the precession angle. In the second, an LGVI-based nonlinear model predictive controller actively drives the RWs to track a continuously scanning spin–precession reference trajectory while satisfying RW torque and momentum constraints. In an accelerated 10-day simulation of a complete 360-deg scan, the NMPC remained convergent throughout all 43,200 control intervals, with a maximum spin-axis tracking error of 0.09084 deg, and completed the scan without RW momentum unloading. A separate parametric study of the initial total-angular-momentum condition further showed that, for the prescribed reference trajectory without momentum unloading, the zero-total-momentum case reduced the transverse RW momentum and peak-torque requirements by approximately 58%. These results demonstrate the effectiveness of combining geometric attitude dynamics, bias-momentum motion generation, and constrained predictive control for long-duration LiteBIRD attitude control.