Speaker
Description
The exponential increase in satellite deployments into Earth’s orbits over the last decade has significantly augmented space environment congestion. This escalation in orbital traffic has simultaneously led to a marked increase in space debris, posing substantial risks to both operational satellites and future space missions. Consequently, addressing the mitigation of space debris has become a critical focus within the fields of Space Surveillance and Tracking (SST) and Space Environment Preservation (SEP). In recent years, several initiatives have emerged to address the issue of space debris, such as Active Debris Removal (ADR) and In-Orbit Servicing (IOS), which aim to mitigate associated risks through targeted removal and maintenance operations in orbit. These activities require precise knowledge of satellite attitude to successfully plan and execute operations for safely approaching, capturing, and manipulating objects in space, thereby enhancing mission success and minimizing the risk of generating additional debris.
The objective of this work is to accurately estimate the attitude of passively rotating Resident Space Objects (RSOs) that could be targeted for debris mitigation measures. This entails addressing the light-curve inversion problem, where the attitude is inferred from photometric intensity measurements of the light reflected from the space object and detected by a ground-based optical sensor. The investigation is specifically centred on inactive space objects with known geometrical shapes. The light-curve inversion problem presents two significant challenges: its highly nonlinear measurement function and the intrinsic ambiguity in measurements, namely the potential existence of multiple attitude solutions for a given set of observations. To address these challenges, this study introduces an advanced attitude inversion method using a particle filter. The particle filter is an approximate Bayesian estimator that represents probability densities using a weighted set of samples (particles), enabling effective handling of multimodal probability density functions (PDFs). Therefore, the particle filter surpasses the limitations associated with other estimation methods, such as the Least Squares Method (LSM) and the Unscented Kalman Filter (UKF), which have already been comprehensively analysed by GMV. The resolution of the attitude inversion problem also requires a high-fidelity light-curve simulator. Photometric observations are simulated using GMV's Grial tool, an advanced high-fidelity simulator implemented in OpenGL that computes the contribution of reflected light from each illuminated and visible pixel on a 3D shape. It employs a Bidirectional Reflectance Distribution Function (BRDF) based on the actual optical parameters of the object and accounts for shading interactions between different parts of the object to ensure realistic simulations.
The presentation delves into the challenges posed by the light-curve inversion problem and explains the novel attitude inversion method based on particle filtering and the aforementioned light-curve simulation tool. The attitude of space objects with various geometries is estimated from both simulated and real light-curves to demonstrate the accuracy of the proposed methodology in determining the true attitude of the space object. The analyses conducted thus far yield excellent results for simulated light-curves, while also demonstrating promising outcomes for real light-curves, thereby indicating the robustness and potential applicability of the method in practical operational scenarios. Various configurations of the particle filter, including sequential and pseudo-batch variants, are also investigated to highlight their advantages and limitations, as well as to evaluate their accuracy and computational performance. Finally, future work is outlined, focusing on further optimizing these methods to enhance their effectiveness across various operational scenarios.
Keywords: space debris removal, photometric measurements, space object characterisation, attitude estimation, particle filtering, light-curve simulation
