RII Track-4 NSF: Robust, Predictive, and Learning Guidance Algorithms for On-Orbit Servicing and Assembly Using Multiple Space Systems

RII Track-4 NSF：使用多个空间系统进行在轨维修和组装的稳健、预测和学习制导算法

• 批准号: 2229756
• 负责人: Hyeongjun Park
• 金额: $ 18.05万
• 依托单位: New Mexico State University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2023
• 资助国家: 美国
• 起止时间: 2023-01-01 至 2023-12-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2229756&HistoricalAwards=false
• 关键词: RII; Track; NSF; Robust; Predictive

Advancements in small space systems enable a variety of on-orbit servicing and assembly missions, such as on-orbit refueling, assembly of complex objects, rescue of damaged space systems, and space debris removal using multiple small space systems. As more space systems will be utilized for sophisticated and critical on-orbit servicing missions, the importance of robust, efficient, and highly accurate autonomous close-proximity maneuvering and docking capability is growing. However, autonomous close-proximity maneuvering is challenging due to uncertainty, disturbances, and multiple satellites’ variable operational constraints, such as actuator saturation, collision avoidance, being within the bounds of the feasible operating region, and achieving a given formation. This project will provide an opportunity for the researcher to collaborate with the Spacecraft Robotics Laboratory at the Naval Postgraduate School. This collaboration will include performing basic and applied research on predictive and learning path-planning algorithms for autonomous close-proximity maneuvering required in on-orbit servicing missions using multiple space systems. This Research Infrastructure Improvement Track-4 EPSCoR Research Fellows (RII Track-4) project provides a fellowship to an Assistant professor at New Mexico State University (NMSU) and support for an NMSU graduate student. The primary goal of this EPSCoR RII Track-4 fellowship project is to develop and test novel robust guidance algorithms based on robust model predictive control and reinforcement learning that can deal with uncertainty and disturbances in formation flying and proximity maneuvering of multiple space systems for on-orbit servicing and assembly missions. Specifically, this project will offer the NMSU team access to world-class facilities for hardware-in-the-loop (HIL) simulation of space proximity operations using multiple space systems to test and validate the proposed algorithms in real-time. Computing the optimal solutions for robust model predictive guidance and control with nonlinear constraints in real time is a significant intellectual challenge. Researchers will develop a rapid optimization framework to address this challenge in solving robust and predictive optimization problems. This research will also contribute to developing a reinforcement learning framework that learns trajectories and control inputs of the predictive approaches. This fellowship project will expand the PI’s research capacity and transform his career path towards a promising direction in guidance algorithm development and in operations using multiple space systems. The outcomes of this project will contribute algorithmically to guidance of on-orbit servicing and assembly missions, and extend the algorithms to a wide range of autonomous applications using multiple space systems. This project will also foster a strong partnership between NMSU, a designated Hispanic-Serving Institution, and the Naval Postgraduate School to increase the participation of students underrepresented in STEM disciplines.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

小型空间系统的进步使各种在轨服务和组装任务成为可能，例如在轨加油、组装复杂物体、救援受损空间系统以及使用多个小型空间系统清除空间碎片。随着越来越多的空间系统将被用于复杂和关键的在轨服务任务，鲁棒、高效和高精度的自主近距离机动和对接能力的重要性正在增长。然而，自主近距离机动是具有挑战性的，由于不确定性，干扰，和多颗卫星的可变操作约束，如致动器饱和，碰撞避免，在可行的操作区域的范围内，并实现给定的编队。该项目将为研究人员提供一个与海军研究生院航天器机器人实验室合作的机会。这项合作将包括对使用多个空间系统的在轨服务任务所需的自主近距离机动的预测和学习路径规划算法进行基础和应用研究。这个研究基础设施改善轨道-4 EPSCoR研究员（RII轨道-4）项目提供奖学金，以助理教授在新墨西哥州州立大学（NMSU）和支持NMSU研究生。这个EPSCoR RII Track-4奖学金项目的主要目标是开发和测试基于鲁棒模型预测控制和强化学习的新型鲁棒制导算法，这些算法可以处理多个空间系统编队飞行和接近机动中的不确定性和干扰，用于在轨服务和组装任务。具体而言，该项目将为NMSU团队提供世界一流的设施，用于使用多个空间系统对空间邻近操作进行硬件在环（HIL）仿真，以实时测试和验证所提出的算法。真实的实时计算非线性约束下鲁棒模型预测制导与控制的最优解是一个重大的智力挑战。研究人员将开发一个快速优化框架，以解决鲁棒性和预测性优化问题。这项研究还将有助于开发一个强化学习框架，该框架可以学习预测方法的轨迹和控制输入。该研究金项目将扩大PI的研究能力，并将他的职业道路转向制导算法开发和使用多个空间系统的操作方面。该项目的成果将有助于在算法上指导在轨服务和组装任务，并将算法扩展到使用多个空间系统的广泛自主应用。该项目还将促进NMSU，一个指定的西班牙裔服务机构，和海军研究生院之间的强有力的伙伴关系，以增加学生在STEM学科代表性不足的参与。这个奖项反映了NSF的法定使命，并已被认为是值得通过使用基金会的智力价值和更广泛的影响审查标准进行评估的支持。