A novel control allocation algorithm based on genetic algorithm (GA) and quadratic programming

Aim: Traditional algorithms are, in our opinion, not satisfactory for bringing the efficient mathematical technique of quadratic programming into the sophisticated problem of allocation of control to a number of effectors. So we propose a novel control allocation algorithm. Section 1 of the full paper briefs the characteristics of multi-effector control allocation. Section 2 explains in some detail the mathematical model of control allocation. The core of section 3 is that the quadratic programming converts the control allocation problem into a constrained optimization one, which can be solved by our GA, and that it presents a four-step procedure for implementing our GA. Section 4 discusses the simulation of a symmetrical flight vehicle; the simulation results, given in Figs. 2 through 5, show preliminarily that: (1) with our novel algorithm, each effector can obtain an appropriate deflection output; (2) if an effector sometimes fails, its control efficiency needs only to be updated to zero; thus our novel algorithm is simple and robust.