执行器不确定系统鲁棒预测动态控制分配策略  

Robust predictive dynamic control allocation for uncertain actuators

作　　者：陈勇[1] 董新民[1] 薛建平[1] 王发威[1] 

机构地区：[1]空军工程大学工程学院,陕西西安710038

出　　处：《控制理论与应用》2012年第4期447-456,共10页Control Theory & Applications

基　　金：航空科学基金资助项目(2008ZC01006)

摘　　要：针对多操纵面级联飞行控制结构中执行器存在多面体不确定的问题,提出了一种基于鲁棒预测控制理论的动态控制分配策略.考虑位置约束和速率约束,建立了多面体不确定冗余执行器的增广控制模型;以执行器状态和虚拟指令跟踪误差为增广变量构造二次型李亚普诺夫函数,将无穷时域Min-Max非线性规划转化为线性矩阵不等式凸优化问题,设计了保守性小的鲁棒预测控制律.各个控制指令汇集到一个混合优化控制分配器,由它分派控制指令,以最优地补偿执行器的不确定动态特性.仿真结果表明,该策略可综合补偿执行器的多面体不确定性,在操纵面偏转范围内精确地跟踪虚拟指令,保证了闭环系统的稳定性,具有较好的鲁棒性.To deal with the polytopic uncertainties of the actuators in the cascaded multi-effector flight control system,we present a dynamic control allocation scheme based on the robust predictive control theory.An augmented control model for the polytopic uncertain redundant actuator is built in considering the position and rate constraints.The actuator state and the virtual command tracking error are taken as the augmented variables in constructing the quadratic Lyapunov function.The ＂Min-Max＂ nonlinear programming in the infinite time domain is converted to a linear matrix inequality convex problem.Thus,we develop an optimal robust predictive controller with less conservativeness.Control commands are integrated into an allocator of mixed optimization,which distributes the control to compensate dynamic uncertainties of actuators.Simulation results show that the proposed scheme compensates the actuator polytopic uncertainties,tracks the virtual commands precisely within the control surface deflections,and guarantees the stability of the closed-loop system with relatively strong robustness.

关 键 词：级联飞控系统 不确定执行器 动态控制分配 鲁棒预测控制 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置] V249.1[自动化与计算机技术—控制科学与工程]