A Fast Quadratic Program Solver for Constrained Control Allocation

Systems with a redundant set of actuators are to be found in many domains which are characterized by high demands on reliability and performance. The ambiguity in the choice of actuator commands caused by redundancy can be resolved by control allocation (CA). One variant of CA involves the formulation of constrained quadratic programs which have to be solved during controller operation. The corresponding method proposed in this work is based on the well-known penalty function method and enables a fast solution of those optimization problems. Two strategies for dealing with infeasible CA problems are suggested. After introducing the concepts extensive randomized tests with comparisons to established solvers demonstrate the effectiveness of the algorithm.