ESA GNC Conference Papers Repository

NASA's baseline Space Launch System (SLS) flight control system (FCS) design includes an adaptive component that modifies the total attitude control system response to provide the classical gain-scheduled control architecture with additional performance and robustness. This algorithm, termed Adaptive Augmenting Control (AAC) was initially developed under the Constellation Program (CxP), analyzed as a side-study for Space Launch System (SLS) Design Analysis Cycle (DAC)-1 (May 2012), and baselined as part of the SLS flight control system (FCS) architecture since DAC-2 (November 2012). The functionally intuitive design was shown to significantly enhance robustness in test cases without negatively impacting performance within the design envelope. The post-Preliminary Design Review (PDR) version of the SLS FCS flight software prototype, including the AAC, was flight tested on a piloted Fighter/Attack (F/A)-18 at NASA Armstrong Flight Research Center. The aircraft acted as a surrogate launch vehicle by mimicking the pitch attitude error dynamics of the more massive, less responsive SLS for the completion of 100+ SLS-like trajectories. Following the aforementioned algorithm development, maturation, and test activities, the NASA Engineering and Safety Center (NESC) and the Space Launch System (SLS) Program performed a comprehensive assessment of the stability and robustness of AAC. This paper will provide an overview of the approach, specific analysis techniques, and outcomes that were particularly relevant for the SLS Program. The standard launch vehicle flight control analyses performed prior to this joint SLS-NESC assessment, as part of SLS analysis cycles, were a combination of (1) frequency-domain stability analysis based on linear theory, and (2) high-fidelity Monte Carlo simulations. The former has shortcomings because it requires the linearization of the nonlinear AAC algorithm. The latter is of limited value because the core control algorithm (without AAC) is able to accommodate the dispersions and ACC is not substantially engaged. These analyses did not reveal any detrimental behavior, but neither did they fully exercise the algorithm. Thus, it was deemed prudent to commission a comprehensive, multifaceted analysis of the stability of the AAC algorithm and it is the overview of these analyses which is the subject of this paper. Multiple techniques that specifically target the SLS AAC were commissioned, with each technique adding its own valuable insights. The following analyses were included: Lyapunov-based stability analysis, classical stability analysis with static AAC gain variations, circle criterion-based analysis of the FCS with a time-varying element, time-domain stability margin assessment, Monte Carlo simulations with expanded dispersions, and an extensive set of stressing cases. Several of the completed analyses focused on determining whether the inclusion of AAC introduced risk to the FCS, while others quantified the benefits of the adaptive augmentation. An overview of the analyses that were applied to the SLS AAC and major findings will be provided in the paper.