Ultralight deployable spacecraft structures are at least an order of magnitude lighter and more volume efficient for launch than state-of-the art spacecraft structures of the same size, thanks to advancements in thin lightweight composites and new packaging schemes. As a result, they are suitable for a wide range of applications requiring extremely large apertures. For example, one of the application that we are currently interested is in-space computer servers. These ultralight spacecraft could not only support photovoltaics that generate large amounts of power, but at the same time act as extremely large radiators for dissipating heat. Similarly, for space-based solar power missions, they can hold functional layers of photovoltaics and phased array antennas to provide uninterrupted energy transmission to any location on Earth. However, this improvement in mass/volume efficiency comes at a price. Ultralight spacecraft exhibit high structural compliance posing new challenges in spacecraft attitude control, which includes residual vibrations in the flexible structure and decreased pointing accuracy, leading to new requirements in structure/control design and testing of these architectures. The present work demonstrates new maneuver techniques and testing methodologies for Caltech’s Space Solar Power Project (SSPP) architecture. Real-world structures are imperfect and require ground tests to characterize their in-space behavior. We present an experimental methodology to perform slew maneuvers to help study the simultaneous behavior of the actuator, control scheme, and the flexible structure response. We present the design of a momentum device prototype and a small-scale, lightweight, flexible structure built for performing slew maneuvers. The flexible structure is designed to be highly compliant along the maneuver direction, allowing the study of its elastic deformation. Smooth polynomial maneuvers are performed with the actuator-structure system mounted on an air bearing; the central hub orientation and flexible structure response is measured. Assuming deformations to be comprised of a single dominant mode, input shaping is applied to the smooth polynomial maneuver, using Zero Velocity (ZV) shaper to suppress vibrations post maneuver. Input-shaping experiments show a remarkable reduction in structural vibrations for high-agility rotations, which directly contributes to an increased power transmission capability in full-scale spacecraft. Although small-scale prototypes are useful for demonstrating control techniques and system-level performance, its quite complex to predict results for full-sized space structures given the lack of detailed similitude models for a combination of actuator-structure behavior and real-world defects. Moreover, full-scale structures are difficult to test on the ground in their deployed configuration, due to limitations in the current gravity compensation techniques. With increases in computational capability, data-centric system identification is the solution to robust modeling of real-world structures. These techniques are powerful not only in decomposing spatial-temporal variations but also in finding complex nonlinear behaviors. In the present work, reduced-order models are created from actuator-flexible structure experiments with equation-based nonlinear extensions to existing data-driven techniques. These models are then used to explore vibration-optimal agile maneuvers without making any assumptions about structural properties. These maneuvers, when tested in the experiments, result in fast rotations with a significant reduction of residual vibrations. This work directly contributes to improving the robustness of existing control schemes, enabling scalability to larger spacecraft and decreasing sensitivity to variations in the deployed structure over time. Together, these results demonstrate that large ultralight spacecraft can be designed and robustly controlled while addressing unwanted effects due to their low stiffness. This development enables their application to next-generation space missions such as space-based solar power, in-space servers, solar sail demonstrations, and much more.
Divesh Soni (Thu,) studied this question.