Model Predictive Control for Sounding Rocket Rate and Attitude Control

The Rate Control System (RCS) is one of the systems responsible for creating the micro-gravity (μg) environment for the experiments on board a sounding rocket. The quality of the μg environment is highly affected by the angular rates of a body in free fall and the results obtained from the experiments are dependent on it. This thesis explores a Model Predictive Control (MPC) approach to improve the control algorithm in contrast to previously used Bang-Bang controllers. The performance of said algorithms is tested in a simulation environment with values obtained from flight data. Based on the sensors used in the current RCS an attitude control is also proposed leveraging the MPC optimal control output.
Based on the results, the proposed algorithm outperforms the currently used techniques when dealing with rate control as well as showing how the Cold Gas System (CGS) can operate with only one working pressure compared to multiple ones, therefore reducing price, complexity, and weight. By using MPC the current system can also provide fast attitude control capabilities without the need to change the sensors used. Given its generality, the method can also be used to create special constraints on rates and maneuvers if the model is expanded with kinematics. By using quaternions, the MPC model remains open to interfacing with state estimation techniques if more sensors were to be added to increase attitude determination accuracy. In conclusion, the proposed method provides rate and attitude optimal controls paving the way for a better μg environment for sounding rockets
experiments.