引用本文:陈勇,董新民,薛建平,郭创.过驱动飞行器线性参变鲁棒预测跟踪控制策略[J].控制理论与应用,2013,30(4):432~442.[点击复制]
CHEN Yong,DONG Xin-min,XUE Jian-ping,GUO Chuang.Robust linear parameter-varying predictive tracking control scheme for over-actuated vehicle[J].Control Theory and Technology,2013,30(4):432~442.[点击复制]
过驱动飞行器线性参变鲁棒预测跟踪控制策略
Robust linear parameter-varying predictive tracking control scheme for over-actuated vehicle
摘要点击 2860  全文点击 1650  投稿时间：2012-05-17  修订日期：2012-11-18
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/CTA.2013.20536
  2013,30(4):432-442
中文关键词  过驱动飞行器  线性变参数  鲁棒预测  控制分配  跟踪控制
英文关键词  over-actuated vehicle  linear parameter-varying  robust prediction  control allocation  tracking control
基金项目  航空科学基金资助项目(20110896009).
作者单位E-mail
陈勇* 空军工程大学 工程学院 cheny_043@163.com 
董新民 空军工程大学 工程学院  
薛建平 空军工程大学 工程学院  
郭创 空军工程大学 工程学院  
中文摘要
      针对飞行包线内过驱动飞行器参数线性时变、控制量多增益调度难的问题, 基于约束动态控制分配提出了一种主环线性参变鲁棒预测跟踪控制器. 采用飞行状态增量以及跟踪指令误差为增广变量, 建立了过驱动飞行器的多面体线性参变离散控制模型. 引入H∞性能指标, 通过参变量转换将无穷时域min-max非线性鲁棒预测控制优化问题约简为线性矩阵不等式凸优化问题. 以操纵面偏转跟踪误差和偏转变化量为混合二次型指标, 建立了包含执行器动态性能的无约束控制分配模型. 进而考虑操纵面的物理约束, 设计了级联动态控制分配策略. 仿真结果表明, 该方法能够生成合理的操纵面控制指令, 实现参考指令的有效跟踪.
英文摘要
      To deal with the time-varying parameters in large envelope of an over-actuated air vehicle with excessive effectors, we propose a method to design the robust predictive tracking controller in the major loop with linear varying parameters based on the constrained dynamic control allocation. The polytopic linear parameter-varying discrete model of over-actuated vehicle is built by adopting the increments of flight states as well as the errors of virtual controls as augmented variables. With the H-infinity objective performance, the infinite time domain ‘min-max’ nonlinear robust predictive optimization problem is converted into the simplified convex linear matrix inequalities. By introducing the tracking errors and varying magnitudes of the effector deflections, we built the unconstrained control allocation model containing the effector dynamics. By considering the physical constraints on effectors, we design the cascaded dynamic control allocation. Simulation results show that this scheme can generate appropriate control instructions for effectively tracking reference commands.