A modified weighted pseudo-inverse control allocation using genetic algorithm

Control allocation is used to distribute the control command into multiple effectors without exceeding the saturations. A powerful control allocation algorithm could enhance the performance of the control system. This paper proposes a new control allocation methodology for the two desired moment problem. The proposed strategy is based on the weighted pseudo-inverse. It has high computational efficiency and easy structure, but could not give the optimal result. Genetic algorithm is used to optimize the weighted matrix for better allocation result. Furthermore, the plane of desired moment is divided into several parts and each part will be bound with one weighted matrix chosen by genetic algorithm. Multiple weighted matrices could guarantee better performance in each part than one constant matrix. To demonstrate the effectiveness of the algorithm, direct allocation and the standard weighted pseudo-inverse are used for comparison. Simulation results show that the newly proposed algorithm could obtain allocation results approximate to or identical with the optimal solutions.