Control allocation for the X-33 using existing and novel quadratic programming techniques

Control allocation (CA) is the distribution of control effort among a set of (possibly redundant) effectors to achieve a desired objective. In the case of aerospace vehicles, CA is the selection of effector commands in order to induce a set of commanded moments (torques) on an aerospace vehicle. These moment commands are typically generated by an autopilot/attitude control law that operates in closed loop with a guidance (attitude command) law as well as a CA law and the vehicle itself. In this paper, several quadratic programming (QP) based CA laws in the context of control for an X-33 based next generation aerospace vehicle model are examined. These laws include least squares (no inequality constraints in QP), direction preserving CA, sign-preserving CA, frequency weighted CA and combinations thereof. These methods are evaluated in the context of both nominal and failed effector behavior so that reconfiguration behavior can be examined. Our simulation results to date indicate that a CA law based on frequency-weighted sign preserving QP problem provides superior performance to the other methods examined.