Laurence Blacketer, Northern Space & Security Ltd.
Keywords: Optical, Light curves, Attitude determination, Synthetic light curve model.
Abstract:
Knowledge of space object attitude state is required for range of different applications. One example is accurate modelling of forces such as solar radiation pressure and atmospheric drag, which are required for the most accurate predictions of an objects motion. Accurate information on space object attitude state could therefore be used to improve the accuracy of conjunction assessment and reentry analysis. Another example is active debris removal (ADR). For an ADR mission to interface with an object, such as with a robotic arm, the object’s attitude state must be known. Additionally, it must be determined if the object’s attitude state is within the capabilities of the ADR spacecraft to interface with and remove, such as due to the angular rotation rate being too high. This therefore also requires the capability predict future attitude states, so that the object suitability can be assessed prior to launch.
In the cases of inactive space objects that are unable to communicate information regarding their attitude state, the state must be determined using external observation data. Brightness verses time measurements collected using ground-based optical telescopes have proven to be a capable and cost-effective method for collecting this observation data. Commonly known as `light curves’, numerous papers have been published that show a strong correlation between variations in the brightness signal and a space object’s attitude state. This work builds on substantial research heritage of using light curves to analyse natural space objects, including properties of their attitude state, such as asteroids.
This paper demonstrates a technique for attitude determination of two classes of space object – rocket bodies and `box-wing’ satellites. A box-wing satellite is defined as a satellite with a box-like bus and two symmetrically placed solar panels. In each case, the initial problem is constrained to inactive objects that exhibit highly periodic brightness variations, and the rotational states are assumed to be rotating about their principle axis of inertia (a `flat-spin’). Angular velocity vectors are then determined using a brute force inverse modelling technique whereby a real light curve is simulated using a synthetic light curve model in combination with a bidirectional reflectance distribution function (BRDF) and an simplified faceted geometry model. The final angular velocity vector solution is selected as the synthetic light curve that best-fits the real light curve data, which is evaluated by measuring the root mean square error between the two signals.
The results show that this technique can be used to determine unique angular velocity vector solutions for both box-wing satellites and rocket bodies. This is achieved using low facet count simplified geometry models, showing that high fidelity geometry models are not necessary. Furthermore a small number of unique BRDF parameters are used for different zones of the geometry models. These factors greatly decrease the computational complexity of the simulations. The future work suggestions explore how the limitations of technique could be evaluated through application to a wider range of scenarios and object types.
This work was in part funded by a United Kingdom (UK) National Space Technology Programme project to explore techniques for attitude determination and prediction of inactive spacecraft in support of a UK ADR mission.
Date of Conference: September 27-20, 2022
Track: Non-Resolved Object Characterization