Towards New V&V in AOCS/GNC for Industrial Efficiency

Abstract

Current and future space missions require increasingly more complex AOCS/GNC designs. Multiple drivers can be identified for this trend, such as larger deployable structures and the extended operational range due to more autonomous spacecraft. For each of these control problems, tailored solutions have to be developed with stringent performance requirements, while staying within time and cost constraints. The drivers highlight the need for efficient analysis tools that can support the Verification and Validation (V&V) process of AOCS/GNC systems. As pointed out in [1], the design phase of a project roughly takes about 20% of the total time, while the remaining 80% are used for planning and executing V&V activities. The increasing system complexity will only make this V&V gap worse. Thus, reducing the V&V effort through enhanced methods is a promising field for improved industrial efficiency. This paper presents the continued work on this topic building on the Airbus contribution to the multi-agency workshop & seminar series [2], [3]. A systematic and detailed framework to assess the robust stability and performance of uncertain systems is proposed, which aims to reduce the V&V engineering effort. Common V&V approaches often rely on large-scale simulation campaigns to cover the various system uncertainties in form of probabilistic Monte Carlo analyses. This process is very time consuming for complex simulations, does not give insights into the uncertainties that drive the stability and performance requirements, and might fail to identify critical scenarios. An alternative is to focus on worst-cases through analytical robustness analysis. Although techniques, such as the μ-analysis, have been available for decades, they have not been integrated into standard industrial processes. To facilitate the use of robustness analysis techniques, a step-by-step approach has been developed that results in an enhanced V&V framework. This framework combines various V&V techniques including worst-case analysis, sensitivity analysis, and optimization within both the time and frequency domain. It consists of 1) modelling the uncertainties through the Linear Fractional Transform (LFT) framework and forming the models from the LFT for stability and performance metrics, 2) defining the robustness metrics under analysis 3) identifying system driving uncertainties through a sensitivity analysis, 4) computing the worst-case degradation of the stability and performance metrics, and 5) performing analysis on the nonlinear simulator. The steps 1) - 4) are all performed within the frequency domain, whereas the last step is computed in the time domain. If the current control design fails the requirements, the obtained worst-cases can be used to synthesize a robust controller. To illustrate the steps in the enhanced V&V framework, an analysis on the robust attitude performance based on MetOp-SG (SAT-B) is presented. The uncertain multi-body dynamics are modelled according to the modular concept in [4] resulting in a compact LFT form. This model is used to analyse the degradation of stability metrics, such as gain, phase, and modulus margin, and performance metrics, such as pointing metrics defined in [5]. Additionally, the uncertainties that are driving these degradations are identified. This information is useful to further improve the control design and can be used as a model reduction technique for the LFT. Finally, the resulting worst-cases from the frequency domain are compared with nonlinear simulations in the time domain. Lastly, conclusions will be drawn how the enhanced V&V framework can potentially improve the current industrial standards. This comparison identifies future work that is necessary to pave the way for increased industrial efficiency in the future. [1] Dennehy, C., Bennani, S., Shankar, U., Vandersteen, J., and VanZwieten, T., “Verification and Validation (V&V) of Guidance & Control Systems: Restults From The First Inter-Agency Workshop [2] Winkler, S., Chapman, P., Juanpere, X. M., Ott, T., “AOCS/GNC Challenges, Solutions and Beyond: Engineering Passion versus for Industrial Efficiency”, Multi-Agency Workshop & Seminar Series, 2021. [3] Martin, M., Belien, F., Winkler, S., „Towards Increased AOCS/GNC Industrial Efficiency: Robust Performance Analysis Considering Real Mission Constraints”, Multi-Agency Workshop & Seminar Series, 2021. [4] Alazard, D., Sanfedino, F., “Satellite dynamics toolbox for preliminary design phase” In : 43rd Annual AAS Guidance and Control Conference, 2020. [5] ESSB-HB-E-003 Working Group, “ESA pointing error engineering handbook ESSB-HB-E-003,” Tech. rep., ESA, 2011.