基于高阶奇异值分解的LPV鲁棒控制器设计

发布时间：2018-05-03 23:02

本文选题：高超声速飞行器 + 飞行包线　；参考：《北京航空航天大学学报》2016年07期

【摘要】：针对高超声速飞行器线性变参数(LPV)模型建模中准确度与复杂度之间的矛盾性,提出了一种基于网格化张量及高阶奇异值分解(HOSVD)的凸多胞LPV模型建模及控制方法。首先基于雅可比线性化给出了大包线网格化模型的张量描述形式,然后提出了一种基于HOSVD的多胞LPV模型生成算法,将网格化模型表述为有限个线性时不变(LTI)顶点及权重函数的组合,并基于舍弃的奇异值给出了建模误差的指标,最后结合某航天飞机再入段六自由度非线性模型进行了大包线鲁棒LPV控制器设计与仿真验证,结果表明该方法可获得计算复杂度低且保证建模精度的LPV模型,设计的鲁棒变增益控制器能够使系统快速跟踪姿态角指令信号,并能够保证系统的稳定性和鲁棒性。
[Abstract]:Aiming at the contradiction between accuracy and complexity in the modeling of hypersonic vehicle with linear variable parameters, a modeling and control method of convex polycell LPV model based on gridded Zhang Liang and high order singular value decomposition (Hos VD) is proposed. In this paper, Zhang Liang's description of large packet line mesh model is given based on Jacobian linearization, and then a polycellular LPV model generation algorithm based on HOSVD is proposed. The gridded model is expressed as a combination of finite linear time-invariant vertices and weight functions, and the index of modeling error is given based on the abandoned singular value. Finally, the design and simulation of a large envelope robust LPV controller based on a nonlinear model with six degrees of freedom in a space shuttle reentry section are carried out. The results show that the proposed method can obtain a LPV model with low computational complexity and high modeling accuracy. The designed robust variable gain controller enables the system to track the signal of attitude angle quickly, and ensures the stability and robustness of the system.
【作者单位】：北京航空航天大学自动化科学与电气工程学院;
【基金】：国家自然科学基金(61273099) 航空科学基金(20135851041)~~
【分类号】：V249.12

【相似文献】