Expo & Sponsorship Sales:  Schedule a Call  or  +44 7773 770269

Reliable Software Development of a Nanosat Attitude Determination and Control System in a Space-Representative Simulator

Author(s):

Mikel Samson, arcsec
Rhimas Van de Putte, arcsec
Wim De Munter, arcsec
Tjorven Delabie, arcsec
Andrew Hyslop, ESA
Enrico Melone, ESA


Presenter:

Mikel Samson, Control Engineer, arcsec


Abstract:

The significant growth of the Nanosatellite market is driven by its core advantages: low cost, fast development cycles, and scalability. However, as the market matures and the (scientific) mission potential evolves, the demands for system reliability increase accordingly. The Attitude Determination and Control System (ADCS) is among the most critical systems of a satellite. It provides the pointing functionality and is critical to achieve the mission goals and to keep the satellite in good health. A challenge for performance validation of this system lies in replicating the space environment in simulations and in on-ground test facilities. Furthermore, the low cost and fast development time often drive an incomplete test-campaign, limited to environmental qualification at unit or satellite level. Additionally, the performance is only assessed during on-orbit commissioning, possibly delaying payload operations in case issues need to be resolved.

To bridge the gap between the fast New Space approach and the traditional Guidance Navigation and Control Verification & Validation (GNC V&V) processes, arcsec, a Belgium-based company, started an ESA General Support Technology Program (GSTP) in mid-2024. This paper describes the development work performed in the first phase of the project, where the flight-software is developed and a model-in-the-loop (MIL) performance test campaign is the main output. The ADCS software is developed in MATLAB/Simulink, and accommodates a discrete number of ADCS configurations with varying number of sensors and actuators. Furthermore, it aligns with the SAVOIR autocoding guidelines from ESA to simplify the generation of the C code that can run on the ADCS processor. The usage of code generation also reduces the cost and development time, and ensures traceability between simulations results, hardware-in-the-loop tests, and flight data. Furthermore, a list of software requirements is composed which flows down from the system and component level requirements. The requirements are tracked in the MATLAB Requirements Toolbox which further strengthens traceability, linking functional and performance requirements to both the simulation and software levels.

The MIL simulations incorporate an external open-source space environment simulator called Basilisk for validating the functionality and performance of the various ADCS configurations and modes. Basilisk is a modular tool which allows to compose the space environment and satellite dynamics by modules which are individually verified by unit tests. Additionally, models for each hardware unit are added based on hands-on experience and measurements to maximize the representativeness.

Finally, the outputs of the simulation are linked to their respective performance characteristics. Where possible they are also compared to on-orbit data originating from the SIMBA and RadCube missions, two 3U ESA IOD/IOV satellites flying an arcsec ADCS. Combining the model-based development approach for the simulator with a validation with on-orbit data results in a representative digital twin of the overall system. Consequently, it can be used to test and derive accurate performance and availability characteristics for a range of ADCS configurations such that the reliability and robustness of the developed ADCS are maximized.

One of the future missions on which the arcsec ADCS will be used is the CubeSpec ESA IOD mission. This mission will demonstrate high-spectral-resolution astronomical spectroscopy and requires accurate pointing and fast slew manoeuvres. Therefore, the ADCS will incorporate two star trackers, six fine Sun sensors, a high-precision gyroscope, three high-precision magnetometers, four reaction wheels and three magnetorquers. The composed simulator enabled to validate if the required pointing accuracy of 100 arcsec (3 sigma) around the cross-boresight axes of the payload is achieved. Furthermore, it was validated if the required slew manoeuvres can be performed within the mission schedule.

Technology: Simulation, Modeling and Automation
Date: May 27, 2025 Time: 5:15 pm - 5:30 pm