Spacecrafts with overactuated and redundant thrusters can work normally once some of them are out of work, which improves the reliability of spacecraft in orbit. In this way, the desired command of controller needs to be dynamically allocated among thrusters. Considering that uncertain factors may appear in forms of dynamics, installation errors, thrust deviations, or failures, this paper proposes a robust control allocation under ellipsoidal uncertainty. This method uses the uncertainty ellipsoid set to describe the uncertainty of the thrusters firstly and establish the thrust allocation robust reference model and then transforms it into a cone optimization model which can be solved as an optimized problem. Finally, this paper adopts the interior-point method for solving the optimization problem. In this way, difficulties of solving the problem caused by parameter uncertainties are avoided effectively. Finally, we take satellite rendezvous and docking as simulation scenarios; it is verified that the cumulative distribution error and maximum error can be reduced by more than 15% when the random error of control efficiency matrix is 5%–20%; also, precision of thruster allocation is improved.
An Approach to QP-based Thrust Allocation considering Inflow
