Digital Twin for In-Space Operations

open_access
Open Access
Author:
Burke, Katelyn
Area of Honors:
Aerospace Engineering
Degree:
Bachelor of Science
Document Type:
Thesis
Thesis Supervisors:
Puneet Singla, Thesis Supervisor
Puneet Singla, Thesis Honors Advisor
Roshan Thomas Eapen, Faculty Reader
Keywords:
Astrodynamics
Digital Twin
Simulation
Satellite
MATLAB
Abstract:
Close proximity operations are of essence for in-space servicing, assembly, and manufacturing (ISAM). The simulation conducted has potential implications in this area and has shown promise in being an accurate tool for an on-board system. In this respect, the relative translational and rotational motions between two space objects are modeled and simulated in MATLAB. The MATLAB simulation results have been exported to the high fidelity physics simulator known as Gazebo and is used to simulate on-board sensors such as camera or lidar. The sensor data gathered from Gazebo simulations has been further processed in MATLAB to estimate the relative motion between two space objects. These estimates can be used to control the space objects to ensure safe close-proximity operations. Creation of this system requires multiple different parts to be completed. First is to solve for the translational and rotational motion of the spacecraft using the initial conditions and the ODE45 function in MATLAB. Then, starting co-simulation between MATLAB and Gazebo using the spacecraft motion simulated in MATLAB and the 3-D model of the spacecraft. This co-simulation is able to prescribe a position for the model at each time step and send back data to MATLAB on the system’s connection. Then, a second model is then added to the simulation and is prescribed an orbit synchronized with the first, but with a larger radius. The timing of the two models must match up precisely, and the orbit was created by manipulating the initial conditions and solving using ODE 45. A third model, a scale model of the Earth, is added to create a more accurate environment. The stereo-vision camera is added to the second model and must stay pointed at the first model, which is completed by rotating the model only 360 degrees throughout the orbit and lining up the starting vision of the model with one of the axes pointing at the Earth model. The stereo vision camera is emulated in Gazebo to image the first spacecraft using the cameras on the second spacecraft. These images are transmitted to MATLAB for further processing. In MATLAB, a program will analyze these images and determine the proximity of the two objects. The program tracks the most dominant features and matches them with their counterparts. The function then uses stereo triangulation in order to determine the distance between the camera’s position and that of the model. The outcome of this project would include a working system that is able to approximate distances between two models in which corrective maneuvers can be calculated and applied. It would also demonstrate the usefulness of this system and showcase its possible applications for ISAM. Several challenges posed by this simulation and its applications include the limited technology available onboard the satellite as well as the communication between different programs used to model and adjust this motion.