Distributed active surface compensation for large space-borne mesh reflectors

Abstract

Deployable mesh reflectors in space are vulnerable to surface deviations induced by attitude adjustments, thermal cycling, and material relaxation, which affect electromagnetic performance stability. To meet the requirements of aerospace missions, a distributed active surface compensation strategy based on model predictive control and driven by micro-electromechanical actuators is proposed. A multilayer distributed subsystem model is developed based on the explicit Newmark-β method combined with dynamic substructure modeling method, enabling the formulation of subsystem-level dynamics tailored for distributed control. For each subsystem, the Newmark-β method is applied to discretize the prediction model with information interaction, and a distributed model predictive control method is developed using cable length adjustment as the driving strategy. By introducing the parametric variational principle, the actuator saturation constraint is reformulated as a linear complementarity problem, enabling efficient active surface error compensation. Numerical simulations show that the proposed method is adaptable to different shape adjustment scenarios and the ability to maintain the surface accuracy under sub-controller failures, demonstrating both stability and fault-tolerance. Furthermore, a closed-loop active surface compensation experimental verification system is also built based on the 3-meter-diameter mesh reflector prototype, and the shape adjustment experiments show a real-time surface error reduction of over 90%. These results confirm the practical feasibility and control effectiveness of the proposed distributed model predictive control approach, achieving real-time error compensation and verifying its applicability for large flexible space mesh structures.