KUbeSat Simulator

As part of my work on the KUbeSat team, I worked with a colleague to create a simulink model of the satellite and its subsystems. Our goal here was to provide a systems-level simulation platform that could be used by the team for various analysis tasks.

The model started out as a way to look at how each system interfaces with the OBC and as a way to develop a model for the satellite state machine. This allowed to software team to work on the logic of how the satellite would operate without having any of the actual hardware. As it evolved we modeled power use and data flow as well as a basic model for each subsystem (ADCS, power, communications, data storage, and OBC). This allowed many other teams to use the model to perform basic analysis such as momentum build-up, battery sizing, duty cycle sizing, and others. The outputs were also used to develop a "quick guide" that was distributed among to team to answer common questions at each stage of satellite development.

The model consisted of two layers (environment and hardware) in order to model the satellite and its response to external stimuli. A third sub-layer was also included in the model to contain various physical constants as well as basic parameters of the satellite such as physical size and mass. Breaking the model into these layers allowed us to separate out important aspects of the model and develop them independently. For instance during the development of the model we changed what orbit propagator we used several times from a basic model to a more complex one. Having the two layers separate allowed us to quickly swap the model out without changing how the systems operate.

The environment layer includes an orbit propogator (2 body and J2), air density model, and solar unit vector (ECEF). This layer then outputted any of the values that would "physically matter" to the satellite such as the sun position, current density, and current state vector (position & velocity).

These values are then sent over to the hardware layer. This layer contains models for each subsystem mainly focusing on power usage. The model for the OBC was done using StateFlow and was a great resource for checking out the state machine logic. This allowed the software team to simulate the satellite in orbit and see what signals they may get in order to develop states and cases for switching states.

Overall this experience really helped me gain a deeper understanding of each of the subsystems involved in a satellite and the interfaces between them. It also helped me gain insight into how to develop a complex model collaboratively in MATLAB/Simulink.