[交通运输]AIAAInterior-Point Algorithms for Control AllocationAIAA_2003.pdf

AIAA Guidance, Navigation, and Control Conference and Exhibit AIAA 2003-5566 11-14 August 2003, Austin, Texas 1 Interior-Point Algorithms for Control Allocation John A. M. Petersen and Marc Bodson Abstract -- The paper considers linear programming formulations of control allocation problems, including those associated with direct allocation and mixed -norm objectives. Three interior-point algorithms, well- 1 suited for the control allocation problems, are described in detail with an emphasis on preferred implementations. The three algorithms are called primal affine-scaling, primal-dual path-following, and predictor-corrector path-following algorithms. The performance of each algorithm is evaluated for computational efficiency and for accuracy using linear models of a C17 transport and of a tailless fighter aircraft. Appropriate choices of stopping tolerances and of other algorithm parameters are studied. Results show that real-time implementation of the algorithms is feasible, without requiring excessive number of computations. Of the three algorithms investiga