Conclusions

Good

• Rate control
• Run-time Plotting Tool
• Quaternion Attitude
  • Theta Multiplier with Quaternion Vector Balancing
  • Decomposing a Quaternion
  • Quaternion to Moment Conversion
• Quaternion-Based
  • PID/SMO State Estimation
  • PID/SMC Control
• Sliding Mode Controller**

Less Good

• Experimental Attitude Control
  • Actuators
  • CSS, Three-Axis Magnetometer

TSatPy

• Quaternion Multiplicative Corrections: Ensures numerical stability in rotating systems
• Adaptive Step Algorithms: Corrects for inconsistent step sizes in simulations or experiments
• Quaternion Decomposition: Decouples rotation and nutation control
• Theta Multiplier with Quaternion Vector Balancing: Linearly scales quaternion attitude measures used in state corrections without violating the unit quaternion restriction
• Euler Axis Based Moments: Utilizes the quaternion's Euler axis and angular displacement to calculate linearly scaled control moments

• Concurrent Estimation/Control Algorithms: Evaluates multiple estimation and control methods
• "Run-time" interface: Enables access and visualization of the internal state of the control system during a test or simulation
• Variable System Clock: Alters the speed of simulations to compress long simulations or inspect state changes is slow motion
• Python: Written in a widely used and expressive language that is easy to learn, comes with "batteries-included", and is viable for high performance applications.
• Modular Design: Additional estimation or control algorithms or system models can be written and plugged into the existing system
• Source Agnostic Control System: Observer-based controller logic can be reused to control a variety of rotating system (simulated and physical)
• MIT License: Software will be released publicly on acceptance of this thesis under as MIT License

Future Work

• Release the source code at https://github.com/MathYourLife under an MIT license.
• Port the tPlot library from MATLAB to Python for enhanced visualizations
• Investigate use of coarse sun sensors for nutation detection.
• Develop a better filtering method such as an extended kalman filter.
• Improve the gradient descent algorithm for parameter tuning.
• Modify the Actuator module to operate pneumatic thrusters and calculate pulse width modulations instead of continuous variable voltages.
• Complete the configuration of the python ADCS "twisted" daemon to open a RESTful API interface to query the system's state.
• Upgrade the three-axis magnetometer nutation to one with a less noisy sensor readings.
• Port the ADCS to a lighter test platform with less friction.
• Investigate with estimator and/or controller scheduling.
• Improve the unit test coverage on the TSatPy library.