基于MDDIFXP二次规划的多操纵面冗余优化  

Multi-control Surfaces Redundancy Optimization Based on MDDIFXP Quadratic Programming

作　　者：毕可军[1] 章卫国[1] 池程芝[1] 朱江乐[1] 

机构地区：[1]西北工业大学自动化学院,西安710072

出　　处：《计算机测量与控制》2013年第4期1063-1066,共4页Computer Measurement &Control

基　　金：航空科学基金项目资助(20100753009)

摘　　要：针对多操纵面飞机存在的气动冗余控制分配问题,提出了一种基于最大方向导数增量(Max Direction Derivative Increment,MDDI)不动点(Fixed-point,FXP)的二次规划优化策略;首先,讨论了基于FXP的二次规划;接着,给出了MDDIFXP算法,通过计算当前迭代点沿各个方向的导数获得MDD,取前一迭代点和当前迭代点在MDD上的迭代增量作为当前迭代点的更新迭代增量;随后,证明了MDDIFXP算法的收敛性;最后,通过对不同控制分配优化方法的仿真验证,表明该改进优化算法MDDIFXP的有效性,以及相比原FXP算法的快速性。For the problem of control allocation of aerodynamic function redundancy for multi-control surfaces aircraft, the quadratic programming optimization scheme based on Max Direction Derivative Increment （MDDI） Fixed-point （FXP） is proposed. First, the quad-ratic programming based on FXP is discussed; Then, MDDIFXP algorithm is given, in which, MDD can be achieved by computing the deriva- tive along all directions of current iteration, and the increments of the former and current iteration are taken as the update for current iteration along the MDD. Moreover, the convergence of MDDIFXP is also proved. Finally, the simulation and comparison of different methods are given. The simulation results show the effectiveness and rapidity of MDDIFXP compared with the original FXP.

关 键 词：多操纵面飞机 控制分配 不动点 优化 伪逆法 改进 

分 类 号：V249.1[航空宇航科学与技术—飞行器设计] TP273[自动化与计算机技术—检测技术与自动化装置]