Robust Control of the Flexible Robotic Arm Based on RBF Neural Networks for Fault Repair in Low Earth Orbit Satellite Constellations

In the field of low-orbit constellation fault repair, traditional methods mainly rely on redundant design and ground control, facing problems such as high redundant design cost and limitations of remote intervention. Therefore, this paper proposes an event-driven robust control strategy for flexible robotic arms in low Earth orbit (LEO) constellation fault repair. Using a Radial Basis Function (RBF) neural network to approximate system nonlinearities, the method improves control performance in complex space environments. An adaptive control law, based on the Lyapunov stability principle, updates only when event-triggered conditions are met, reducing computational load. Simulations validate its effectiveness in handling nonlinear dynamics and disturbances, demonstrating potential for precise on-orbit repair tasks such as satellite capture and component replacement, thus improving LEO constellation reliability and lifespan.