Fault-tolerant Satellite Trajectory Optimization and Control using Advanced Nonlinear Predictive Controls

Satellite trajectory control and optimization are complicated due to several factors: nonlinear dynamics with time delays, modeling and parameter uncertainties; flexible modes due to fuel slosh and appendages; constraints on propulsive force and torque inputs; and constraints on acceptable angular rates and attitudes. Existing control approaches based on LTI frequency domain methods trade simplicity and ease of analysis at the expense of performance and efficiency, which translates to increased fuel payload and power requirements. Model Predictive Control (MPC) is an optimal control approach based directly on system models and on-line optimization to follow a desired trajectory, subject to control, safety and mission constraints. Scientific Systems has successfully applied nonlinear MPC to simplified spacecraft attitude control problems. The proposed Phase I effort will demonstrate the feasibility of using advanced Nonlinear MPC for satellite trajectory control/optimization, and quantify the advantages of using NMPC in terms of improved accuracy, reduced propellant usage, and overall mission success. Satellite trajectory and attitude control problems will be formulated for missions of interest to the AFSPC, e.g. orbit insertion, precision pointing/slewing, orbit correction, raising or repositioning. Algorithms will be developed and tested on problems using satellite and planetary models, simulation requirements and mission/task specifics supplied by AFSPC. Prof. John Wen (RPI) will provide consulting services for problem definition, theory and algorithm development of Nonlinear MPC for satellite control.