FindAPhD Weekly PhD Newsletter | JOIN NOW FindAPhD Weekly PhD Newsletter | JOIN NOW

Robust Model Predictive Attitude Control for Multi-Spacecraft Cargo Missions


   School of Aerospace, Transport and Manufacturing (SATM)

This project is no longer listed on FindAPhD.com and may not be available.

Click here to search FindAPhD.com for PhD studentship opportunities
  Dr Leonard Felicetti  No more applications being accepted  Funded PhD Project (European/UK Students Only)

About the Project

Applications are invited for a 3 years PhD studentship in Space Engineering, at Cranfield University, in the field of Spacecraft Attitude Dynamics and Control. The project aims to develop and investigate modern and robust control strategies (e.g. based on model predictive control, on AI etc.) for the attitude control of multi-spacecraft cargo missions and other new space age mission concepts.

New mission concepts and services in the New Space age require a sensible enhancement of the Guidance Navigation and Control (GNC) capabilities. Such a new systems are increasingly required to perform autonomous decision onboard and run complex operations in non-defined scenarios or conditions. In most cases, such algorithms embed model predictive schemes to optimize maneuvers in real-time by offering a wider versatility compared to classical schemes and the capabilities of identifying potential future behaviors and risks for the mission. The implementation of such algorithms becomes challenging when the real-time runs need to be performed with complex and non-linear dynamics, often characterizing the attitude dynamics of modern space systems when fast slew and reorientation maneuvers are requested. Optimization techniques based on convex optimization are generally used to overcome these problems, but part of the actual capability appears yet unexplored and needs further investigation. The problem becomes even more challenging when noisy inputs from sensors need to be considered and fed into the GNC loop, and systems are dynamically changing their mass properties during the mission, e.g., due to deployment of payloads, sloshing effects, and flexibility of some appendages.

The present PhD research aims to investigate the real-time implementation of model predictive attitude control for such kinds of applications. Particular focus will be given to the identification of suitable convex, statistical, linear and non-linear optimization approaches that allow for a real-time implementation under a model-predictive framework for solving attitude dynamics and control problems when unknown or variable mass properties are characterizing the mission. Potential applications will be explored to identify the key drivers and specific performance of the adoption of such schemes in a wide range of possible scenarios. Specifically, the problem of attitude control of a cargo platform for deployment of multi-payload spacecraft will be considered as principal scenario, where the uncertainty and variability of the mass properties of the spacecraft, such as overall mass distribution and moments of inertia and center of mass position, are foreseen as critical elements that should considered since the preliminary phases of the mission. The selected algorithms will be then tested in both simulated and experimental setups to evaluate the actual benefits, compared to other classical techniques (such as Lyapunov-based, H-infinity with mu-synthesis or sliding control techniques), and the potential critical issues that might limit their utilization in a space-based embedded system. Possible other applications, such as cooperative and noncooperative rendezvous, relative GNC for on-orbit servicing, assembly and manufacturing or autonomous pointing during super-fast fly-bys, non-Keplerian orbits and aero-breaking situations will be also explored within this research. A focus should also be put on the implementation of such algorithms in flight hardware, which should be radiation hardened and capable of running on-board optimization in real-time. The acquisition of such specific applied know-how and the development of ad-hoc GNC solutions for attitude control systems represent tangible outcomes of such research that will feed on the current activities of Coactum Space. The partnership will allow the student to develop key software and interact with real under-development projects, get specific mission requirements for the Coactum missions, and access, deploy, and test algorithms in dedicated hardware and systems for the Coactum missions

Applicants shall have demonstrated experience in space dynamics, attitude and orbit control systems, estimation theory and Kalman filters, and classical and modern control techniques, with particular focus on model predictive control strategies and convex optimization. Experience in computer programming (e.g. Python, C++) and relevant engineering software (e.g., MATLAB/Simulink, STK, GMAT, ROS/Gazebo) is an essential asset. Hand-in experience with spacecraft onboard computers or embedded systems such as Nvidia Jetson and relevant AOCS hardware is preferred. Applicants should demonstrate proficiency in technical writing, and a track record with publications in well-recognized journals and conferences is preferred. 

The successful PhD candidate will be based at Cranfield University with the possibility of visiting and collaborating with Coactum SA. Within this project, the candidate will build a strong background in spacecraft attitude dynamics and control as well as advanced control techniques for space applications. Upon successful completion of the PhD, the candidate will be able to carry out independent research and development activities in research centres and companies.


Funding Notes

Sponsored by Coactum SA and Cranfield University, this studentship will provide a bursary of up to £18,000 (tax free) plus fees* for three years.
Search Suggestions
Search suggestions

Based on your current searches we recommend the following search filters.

PhD saved successfully
View saved PhDs