Combined feedback linearization and constrained model predictive control for entry flight

The advantages of constrained model predictive control over proportional integral derivative (PID) control applied to a feedback-linearized entry flight-control problem are discussed. The feedback linearization is based on the full rotational equations of motion rather than on a conventional model derived from time-scale separation. Input and state constraints are applied to avoid input saturations, to guarantee a minimum level of tracking performance,and to avoid physical vehicle state constraints violation. A constraint mapping algorithm is developed to map the input and state constraints on the new inputs after feedback linearization. The performance of the constrained model predictive control design is compared with that of two PID control designs. Simulation of a complete entry flight demonstrates the advantages of the constrained model predictive control design, because control surfaces do not saturate, control actions are more smooth and efficient, no gain scheduling is required, and all performance requirements are satisfied.