Tracking Mode for Attitude Determination

Marissa Myhre, B.Eng. Student 3^rd Year

Supervisor: Regina Lee

Lassonde Undergraduate Research Conference

August 16^th 2023

​

Introduction

2

RSONAR II

• Resident Space Object Near-space Astrometric Research II (RSONAR II)
• Technology demonstration of low-resolution, wide field of view dual-purpose CubeSat compatible star tracker concept for Space Situational Awareness (SSA)
• 4U cubesat with 4 independent systems
• Second iteration of the payload to be flown on a stratospheric balloon introducing attitude determination, RSO detection, and downlink capability

​

3

RSONAR Launch 22 August 2022 Timmins, Ontario   

RSONAR II payload

RSONAR II payload after gondola integration

​

Payload Overview

4

STARDUST Dual-Purpose Payload

• Star Tracker for Attitude and Resident space object (RSO) Detection for Unified Space Technologies
• Real-time attitude determination and RSO detection using a wide FOV star tracker-like camera
• Raspberry Pi 4 Model B
• 8 GB RAM
• 512 GB U3 microSD card
• IDS UI-3370CP-M-GL Rev.2 camera
• Monochrome
• 1” sensor, 5.5 um, 2048 x 2048 pixels
• Paired with 16mm Telephoto lens
• 40 x 40 deg FOV

​

5

https://cdn11.bigcommerce.com/s-am5zt8xfow/images/stencil/1280x1280/products/2144/5829/782_1__82433.1588265316.jpg?c=2

https://canada.newark.com/productimages/large/en_US/02AH3161-40.jpg

https://www.ids-imaging.us/store_us/media/catalog/product/cache/2/image/800x800/9df78eab33525d08d6e5fb8d27136e95/c/a/camera-usb3-ueye-cp-rev2-1.jpg

Attitude Determination

• Attitude = Orientation
• Yaw Pitch Roll Conventions
• ZYX, RHR, intrinsic, and active Euler rotation
• Expressed in ECI frame (Earth Centered Inertial)

6

https://upload.wikimedia.org/wikipedia/commons/3/32/Earth_Centered_Inertial_Coordinate_System.png

https://tensortech.com.tw/how-to-pick-up-a-suitable-cubesat-adcs-attitude-determination-and-control-system-components-or-an-integrated-system-for-your-mission/

Star Tracker (Star Sensor)

• Optical Sensor
• Considerably the most accurate sensor of attitude determination with arcseconds of accuracy
• Lost in Space Mode
• High accuracy low update rate
• Tracking Mode
• Provides a high update rate

​

​

7

Yes

No

Lost In Space Mode Gabriel Chianelli, MSc Student

• Lost in Space Mode integrated by Gabriel Chianelli, MSc Student
• Real-time attitude determination using a wide FOV star tracker
• Lost-in-Space Mode, output quaternion in Earth Centered Inertial (ECI) Frame with < 10 arcmin accuracy (σ)
• Star detection and Center of Mass (COM) centroiding
• Planar area and moment star identification
• QUEST attitude determination

​

8

Tracking Mode

• Capabilities for high update rate attitude determination
• Updates attitude based off of star pixel location displacement
• Tracking mode offers a higher update rate without the need to re-identify the stars

​

• For RSONAR II Tracking mode
• Low-resolution and high field of view camera
• Increasing attitude update rate

​

​

​

9

Objectives

10

Objectives

• Create Tracking mode to complement lost in space mode for attitude determination subsystem
• Achieve a high update rate attitude determination
• Tracking mode for a low-resolution and high field of view camera
• Maintain less than 1 degree pointing error while in tracking mode
• Real-time attitude determination processing

​

​

​

​

11

Methodology

12

Methodology

1. Detect stars
1. Find star pixel locations
2. Compare with previous star locations
3. update the body vector
2. Determine angular rate
3. Update attitude

Conventions:�Attitude data expressed in degrees for a ZYX rotation in the ECI frame with intrinsic, active, and RHR rotation conventions

​

13

Methodology

14

Conventions:

Attitude data expressed in degrees for a ZYX rotation in the ECI frame with intrinsic, active, and RHR rotation conventions

​

Detect Stars Pixel Location

• An image in array format is inputted
• Histogram equalization or a Gaussian Blur is applied
• Gamma correction applied to increase brightness of stars
• If a contort is larger then a threshold of 5px it is assumed to be a star
• The centroid of each star is then recorded each star is then recorded

15

Recorded Stars are boxed in green

Compare With Previous

• Defined radius of which a star can be within for it to be assumed the same star as the previous image
• Detected stars are matched to their previous pixel locations

​

16

Body Vector

17

Body Vector

18

Determine Angular Rate

19

Update Attitude

20

Update Attitude

21

Results

22

Truth Data

• Images from RSONAR were post processed with the open source software Nova Astrometry Net
• Yaw Pitch and Roll in ECI for the entire mission was saved

​

​

​

23

Testing

​

​

24

• Tested on images from RSONAR
• Truth Data is acquired from Astrometry.net’s open source software

STARDUST Field Campaign

• STARDUST ran their field campaign Friday August 4^th 2023
• The campaigns purpose was to test the how the 3 algorithms worked together after being integrated onto the hardware
• Tracking Mode
• Lost In Space (Gabriel Chianelli, MSc Student)
• RSO detection (Perushan Kunalakantha, MSc Student)

​

​

​

25

Images from RSONAR II/ STARDUST field campaign

Expected Outcomes/ Impact

• Fly’s as a part of the attitude determination system on RSONAR II Sub-Payload STARDUST
• To provide attitude updates in ECI Frame with < 1 deg for 5 minute intervals before switching Lost in Space mode back on.
• Expected update rate of 10 Hz
• Demonstrate Star trackers feasible to use on for nano-satellite space missions

​

RSONAR II will be flying from Timmins, Ontario with the current launch window August 13^th – 21^st

​

​

26

RSONAR II in pelican case

Future Work

• Implement Extended Kalman Filter (EKF)
• To minimize the error build up
• Implement variable search Radius
• Based on the angular velocity estimate determine a new search radius for the star in the next image
• Integrate Tracking mode to work with additional attitude sensors
• Sun sensor, horizon sensor, gyros, magnetometers

​

27

Acknowledgements

28

Gabriel Chianelli, MSc Student

Dr. Regina Lee, Supervisor

Perushan Kunalakantha, MSc Student