2.845

2023影响因子

(CJCR)

  • 中文核心
  • EI
  • 中国科技核心
  • Scopus
  • CSCD
  • 英国科学文摘

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

SGCMGs驱动的挠性航天器有限时间自适应鲁棒控制

王璐 郭毓 吴益飞

王璐, 郭毓, 吴益飞. SGCMGs驱动的挠性航天器有限时间自适应鲁棒控制.自动化学报, 2021, 47(3): 641-651 doi: 10.16383/j.aas.c180446
引用本文: 王璐, 郭毓, 吴益飞. SGCMGs驱动的挠性航天器有限时间自适应鲁棒控制.自动化学报, 2021, 47(3): 641-651 doi: 10.16383/j.aas.c180446
Wang Lu, Guo Yu, Wu Yi-Fei. Finite-time adaptive robust control for SGCMGs-based flexible spacecraft. Acta Automatica Sinica, 2021, 47(3): 641-651 doi: 10.16383/j.aas.c180446
Citation: Wang Lu, Guo Yu, Wu Yi-Fei. Finite-time adaptive robust control for SGCMGs-based flexible spacecraft. Acta Automatica Sinica, 2021, 47(3): 641-651 doi: 10.16383/j.aas.c180446

SGCMGs驱动的挠性航天器有限时间自适应鲁棒控制

doi: 10.16383/j.aas.c180446
基金项目: 

国家自然科学基金 61973167

国家自然科学基金 61773211

国家自然科学基金 61673214

国家自然科学基金 61673219

详细信息
    作者简介:

    王璐  南京理工大学自动化学院博士. 主要研究方向为挠性航天器姿态控制. E-mail: timwang21@126.com

    吴益飞  南京理工大学自动化学院副教授. 主要研究方向为机器人控制, 伺服系统控制. E-mail: wuyifei0911@163.com

    通讯作者:

    郭毓  南京理工大学自动化学院教授. 主要研究方向为机器人控制, 航天器姿态控制. 本文通信作者. E-mail: guoyu@njust.edu.cn

Finite-time Adaptive Robust Control for SGCMGs-based Flexible Spacecraft

Funds: 

National Natural Science Foundation of China 61973167

National Natural Science Foundation of China 61773211

National Natural Science Foundation of China 61673214

National Natural Science Foundation of China 61673219

More Information
    Author Bio:

    WANG Lu  Ph. D. at the Automation School, Nanjing University of Science and Technology. His main research interest is flexible spacecraft attitude control

    WU Yi-Fei  Associate professor at the Automation School, Nanjing University of Science and Technology. His research interest covers robot control and servo system control

    Corresponding author: GUO Yu  Professor at the Automation School, Nanjing University of Science and Technology. Her research interest covers robot control and spacecraft attitude control. Corresponding author of this paper
  • 摘要: 针对挠性航天器系统中同时存在单框架控制力矩陀螺群(Single gimbaled control moment gyroscopes, SGCMGs) 摩擦非线性、电磁干扰力矩、惯量摄动以及外部干扰等问题, 提出了一种有限时间自适应鲁棒控制(Finite-time adaptive robust control, FTARC) 方法. 针对系统中存在未知参数的情况, 分别设计自适应更新律, 使得控制器的设计不依赖参数信息, 同时减小外部干扰对系统的不利影响. 应用Lyapunov稳定性理论证明了闭环系统姿态角误差和姿态角速度误差可在有限时间内收敛到原点附近的邻域内. 仿真结果表明, 所提控制律可实现挠性航天器姿态快速机动, 并获得甚高指向精度.
    Recommended by Associate Editor NI Mao-Lin
  • 单框架控制力矩陀螺群(Single gimbaled control moment gyroscopes, SGCMGs) 由于其质量轻、输出力矩大、功耗低等优点, 作为执行机构已广泛应用于航天器姿态控制中. 然而, 尽管SGCMGs等空间设备的组装和测试过程都很精密和规范, 在执行姿态机动任务时, SGCMGs不可避免地会受到一些非线性干扰力矩的影响, 如SGCMGs轴承摩擦和电磁干扰力矩等. 这会使执行机构输出力矩精度受到影响进而影响航天器姿态控制性能. 此外, 由于挠性航天器通常采用中心刚体加挠性附件的结构, 刚体与挠性附件的强耦合效应会影响航天器的姿态控制性能. 航天器燃料的消耗和附件的伸缩等也会引起惯量的变化, 同时航天器系统还会受到外部干扰力矩的影响. 这些因素均对航天器控制性能的提升带来一定难度. 因此, 需要设计鲁棒控制律, 减小航天器和执行机构的不确定性及外部干扰等问题对系统的影响, 满足挠性航天器高性能姿态控制的要求.

    近年来, 国内外学者针对挠性航天器姿态控制系统中惯量摄动、刚挠耦合及外部干扰等问题开展了大量研究工作[1-4], 取得很大进展. 然而, 已有的研究大多假设所设计的控制器输出直接作用于航天器对象上. 在控制精度要求不高的情况下, 这些方法能够满足一定的控制需求, 但在更高精度控制目标下仍具有一定局限性. 为了进一步提高姿态控制精度, 对执行机构中固有的摩擦非线性和扰动力矩等因素的研究越来越受重视. 针对SGCMGs所受非线性摩擦力矩及干扰力矩等问题, 文献[5-8] 从SGCMG框架伺服系统控制角度, 对其所受干扰力矩进行建模与分析, 并设计扰动力矩观测器及终端滑模变结构控制器等, 实现框架伺服系统的高精度跟踪控制, 提高系统的力矩输出精度. 文献[9] 研究了基于LuGre模型的摩擦非线性对航天器姿态控制的影响. 文献[10] 设计了一种非线性自适应姿态控制器来估计和补偿SGCMGs中的动摩擦环节. 文献[11]设计了一种自适应控制器来减小航天器惯量、变速控制力矩陀螺框架和转子惯量不确定性问题对系统的影响. 文献[12] 通过对基于LuGre模型的SGCMGs摩擦非线性力矩进行分析, 提出了一种鲁棒自适应姿态控制律, 减小了摩擦非线性力矩对系统的影响并进一步提高了姿态控制精度. 文献[13-15]针对反作用飞轮中的摩擦非线性问题, 设计了鲁棒自适应控制器及有限时间角速度观测器, 对摩擦非线性进行估计和补偿, 为SGCMGs中摩擦问题的研究提供了借鉴. 针对SGCMGs中电磁扰动力矩对系统的影响, 文献[16] 建立了框架角谐波正弦函数形式的电磁扰动力矩模型, 并针对变周期干扰力矩提出一种干扰抑制滤波器. 电磁扰动力矩会严重影响SGCMGs力矩输出精度, 给航天器姿态机动控制带来了巨大的挑战, 针对以上形式的电磁扰动力矩及航天器惯量不确定性等问题, 文献[17]提出了一种自适应神经网络控制器, 然而在SGCMGs动力学中并未考虑SGCMGs摩擦非线性力矩对航天器系统的影响. 由于SGCMGs力矩输出精度与航天器姿态机动精度密切相关, 因此全面地考虑SGCMGs动力学中存在的非线性问题是十分必要的.

    本文主要针对SGCMGs电磁扰动力矩对航天器姿态控制系统影响进行研究, 同时为减小航天器惯量不确定性、刚挠耦合、及外部干扰上界未知等对系统的不利影响, 为进一步提高系统误差收敛的快速性, 设计一种基于终端滑模的有限时间自适应鲁棒控制(Finite-time adaptive robust control, FTARC) 算法. 基于Lyapunov理论证明了航天器姿态角误差和姿态角速度误差在有限时间内收敛到原点附近的邻域内.

    采用四元数法描述的航天器姿态运动学方程为[18]

    $$ \begin{align} \dot{\pmb{{q}}} = \frac{1}{2} \left[ {\begin{array}{*{20}{c}}q_4{\bf{I_{3}}}+{{q}}_v^{\times}\\-\pmb{{q}}_v^{\rm T}\end{array}} \right]\boldsymbol{\omega} \end{align} $$ (1)

    其中, $ \pmb{{q}} = [\pmb{{q}}^{\rm T}_v\; \; q_4]^{\rm T} $, $ \pmb{{q}}_v = [q_1\; \; q_2\; \; q_3]^{\rm T} $, 且$ {{q}}_v^{\times} $是$ \pmb{{q}}_v $的斜对称矩阵, 即

    $$ \begin{align} {{q}}_v^{\times} = \left[ {\begin{array}{*{20}{c}} {0}&{-q_3}&{q_2}\\ {q_3}&{0}&{-q_1}\\ {-q_2}&{q_1}&{0} \end{array}} \right] \end{align} $$ (2)

    $ \boldsymbol{\omega}\in\textbf{R}^{{3}\times{1}} $为航天器本体坐标系相对于惯性坐标系的角速度矢量. 期望姿态四元数和姿态角速度分别为$ \pmb{{q}}_d $和$ \boldsymbol{\omega}_d $, 定义姿态误差四元数$ \pmb{{e}} = \pmb{{q}}_d^{-1}\otimes\pmb{{q}} = [\pmb{{q}}_{ev}^{\rm T}\; \; q_{e4}]^{\rm T} $.

    $$ \begin{align} \begin{cases} \dot{\pmb{{q}}}_{ev} = \frac{1}{2}[q_{e4}{\bf{I}}_{3}+{{q}}_{ev}^{\times}]\boldsymbol{\omega}_e \\ \dot{q}_{e4} = -\frac{1}{2}\pmb{{q}}_{ev}^{\rm T}\boldsymbol{\omega}_{e} \end{cases} \end{align} $$ (3)

    其中, $ \boldsymbol{\omega}_e = \boldsymbol{\omega}-\boldsymbol{\omega}_r $为角速度误差, $ \boldsymbol{\omega}_r = \widetilde{{{R}}}\boldsymbol{\omega}_d $, $ \widetilde{{{R}}} $为基于四元数描述的姿态变换矩阵.

    SGCMGs驱动的挠性航天器姿态动力学方程和挠性附件振动方程分别为[18]

    $$ \begin{align} & {{J}}\dot{\boldsymbol{\omega}}+{\omega}^{\times}{{J}}\boldsymbol{\omega}+{{C}}^{\rm T}\ddot{\boldsymbol{\eta}} = {\pmb{T}}_{\rm cmg}+{\pmb{T}}_d \end{align} $$ (4)
    $$ \begin{align} & \ddot{\boldsymbol{\eta}}+2{\zeta}{\varLambda}\dot{\boldsymbol{\eta}}+{\varLambda}^2\boldsymbol{\eta} = -{{C}}\dot{\boldsymbol{\omega}} \end{align} $$ (5)

    其中, $ {{J}}\in\textbf{R}^{{3}\times{3}} $为正定对称的转动惯量矩阵, $ \dot{\boldsymbol{\omega}} $为航天器角加速度, $ {\omega}^{\times} $为$ \boldsymbol{\omega} $斜对称阵; $ {{C}}\in\textbf{R}^{n\times{3}} $为航天器本体与挠性附件的耦合阵;$ \boldsymbol{\eta}\in\textbf{R}^{n\times{1}} $为挠性附件的模态坐标; $ {\pmb{T}}_d\in\textbf{R}^{{3}\times{1}} $为空间环境的干扰力矩; $ {\mit\zeta}\in\textbf{R}^{n\times n} $为挠性附件的阻尼比矩阵; $ {\varLambda}\in\textbf{R}^{n\times n} $为挠性附件的模态频率矩阵. SGCMGs的输出力矩$ {\pmb{T}}_{\rm cmg} $ $ \in $ $ \textbf{R}^{{3}\times{1}} $作为航天器姿态机动的控制力矩.

    本文采用金字塔构型的SGCMGs陀螺群作为挠性航天器执行机构. SGCMGs中存在诸多非线性因素, 其中影响力矩输出的扰动力矩主要为摩擦力矩及脉动力矩. SGCMGs动力学模型为

    $$ \begin{align} {\pmb{T}}_{\rm cmg} = -\dot{\boldsymbol{h}}_{\rm cmg}-{\omega}^{\times}\boldsymbol{h}_{\rm cmg}-{A}{\pmb{F}}_f+{A}{\pmb{F}}_d \end{align} $$ (6)

    其中, $ \boldsymbol{h}_{\rm cmg} $为SGCMGs角动量; $ \dot{\boldsymbol{h}}_{\rm cmg} $为框架角速度; $ \dot{\boldsymbol{h}}_{\rm cmg} = h_a{{A}}\dot{\boldsymbol{\delta}} $为角动量变化率, $ h_a $为各陀螺单位角动量幅值; $ {{A}}\in\textbf{R}^{{3}\times{4}} $为雅可比矩阵, 其表达式如式(7)所示[19].

    $$ \begin{align} {{A}} = \left[ {\begin{array}{*{20}{c}} {-\cos\delta_1\cos\beta}&{\sin\delta_2}&{\cos\delta_3\cos\beta}&{-\sin\delta_4}\\ {-\sin\delta_1}&{-\cos\delta_2\cos\beta}&{\sin\delta_3}&{\cos\delta_4\cos\beta}\\ {\cos\delta_1\sin\beta}&{\cos\delta_2\sin\beta}&{\cos\delta_3\sin\beta}&{\cos\delta_4\sin\beta} \end{array}} \right] \end{align} $$ (7)

    倾斜角$ \beta $为金字塔斜面与底面的夹角, $ \boldsymbol{\delta} = [\delta_1 $  $ \delta_2 $  $ \delta_3 $  $ \delta_4]^{\rm T}\in\textbf{R}^{{4}\times{1}} $, $ \delta_i $ $ (i = 1, 2, 3, 4) $为SGCMG框架角.

    注1. 由式(7)可得, $ \|{{A}}\|_{i\infty}<A_m $, 其中, $ A_m $为已知正数且$ \|\cdot\|_{i\infty} $表示矩阵诱导无穷范数.

    SGCMGs摩擦力矩$ {\pmb{F}}_f = [F_{f1}\ \, F_{f2}\ \, F_{f3}\ \, F_{f4}]^{\rm T}\in\textbf{R}^{{4}\times{1}} $与框架角速度相关, 采用LuGre动态摩擦模型的数学描述[20-21]

    $$ \begin{align} \begin{cases} F_i = \sigma_0z_i+\sigma_1\dot{z_i}+\sigma_2\dot{\delta_i} \\ \dot{z_i} = \dot{\delta_i}-\dfrac{|\dot{\delta_i}|z_i}{g(\dot{\delta_i})} \\ \sigma_0g(\dot{\delta_i}) = F_c+(F_s-F_c){\rm e}^{-\left(\frac{\dot{\delta_i}}{V_s}\right)^2} \end{cases} \end{align} $$ (8)

    其中, $ \sigma_0 $为鬃毛刚度系数, $ \sigma_1 $为鬃毛阻尼系数, $ \sigma_2 $为粘性摩擦系数, 三个系数未知. $ z_i $为鬃毛平均变形, $ \dot{\delta_i} $为框架角速度, $ g(\dot{\delta_i}) $为摩擦效应. $ F_c $为库仑摩擦力矩, $ F_s $为最大静摩擦力矩, $ V_s $为Stribeck特征速度.

    单框架控制力矩陀螺中的电磁扰动力矩也是影响力矩输出的不利因素. 该干扰力矩主要来自于扭矩装置框架角速度测速计, 其形式与传感器干扰相似. 尽管现在框架电速计等空间速度装置需要精心手工安装和调试, 这种电磁扰动力矩在扭矩装置的内部仍难以完全消除. SGCMGs扰动力矩中的脉动力矩$ {\pmb{F}}_d = [F_{d1}\ \, F_{d2}\ \, F_{d3}\ \, F_{d4}]^{\rm T}\in\textbf{R}^{{4}\times{1}} $可表示为框架角谐波的正弦函数形式, 与框架角速度和框架角均相关, 表示为[16-17]

    $$ \begin{align} {\pmb{F}}_d = {{K}}_G{{E}}_d\dot{\boldsymbol{\delta}} \end{align} $$ (9)

    其中, 框架前馈环路增益$ {{K}}_G $未知, $ {{E}}_d = $ $ {{\rm{diag}}}\{{E}_{d1}(\delta_1)\ \, {E}_{d2}(\delta_2)\ \, {E}_{d3}(\delta_3)\ \, {E}_{d4}(\delta_4)\} $为4个框架轴的干扰电压矩阵, 其中, $ {E}_{di}(\delta_i) $, $ i = 1, 2, 3, 4 $, 形式为[18]

    $$ \begin{align} {E}_{di}(\delta_i) = \sum\limits_{n = 1}^{10}\left\{\frac{1}{n} \sin(n\delta_i)+\frac{1}{n+1}\cos(n\delta_i)\right\} \end{align} $$ (10)

    进一步整理, 式(6)可写为

    $$ \begin{align} {\pmb{T}}_{\rm cmg} = -(h_a{{A}}+{A}{{F}}_\sigma)\dot{\boldsymbol{\delta}}-{\omega}^{\times}\boldsymbol{h}_{\rm cmg}-{A}{\pmb{f}}+{A}{\pmb{F}}_d \end{align} $$ (11)

    其中, $ {\pmb{f}} = [f_1\ \, f_2\ \, f_3\ \, f_4]^{{\rm T}} $, $ f_i = \sigma_0z_i-\sigma_1|\dot{\delta_i}|z_i/g(\dot{\delta_i}) $, $ i $ $ = 1, 2, 3, 4 $且$ {{F}}_{\sigma} = \sigma{\bf I}_{4} $, $ \sigma = \sigma_1+\sigma_2 $. 由于摩擦力矩有界[20], $ |f_i|\leq{f_m} $, $ f_i $为未知正数. 由注1可得

    $$ \begin{align} \|{A}{\pmb{f}}\|\leq{\|{{A}}\|_{i\infty}\|{\pmb{f}}\|\leq{2 A_mf_m}} \end{align} $$ (12)

    针对挠性航天器存在SGCMGs扰动力矩、惯量摄动及外部时变干扰的情况, 为实现高精度快速姿态机动, 设计有限时间自适应鲁棒姿态控制器, 系统结构如图 1所示.

    图 1  SGCMGs驱动的挠性航天器姿态控制系统结构图
    Fig. 1  Structure diagram of attitude control system for SGCMGs-based flexible spacecraft

    选取辅助变量为

    $$ \begin{align} \pmb{{r}} = \boldsymbol{\omega}_{e}+k_1\pmb{{q}}_{ev}+k_2{\pmb{f}}(\pmb{{q}}_{ev}, l) \end{align} $$ (13)

    快速非奇异终端滑模面为$ \boldsymbol{\omega}_{e}+k_1\pmb{{q}}_{ev}+k_2{\pmb{f}}(\pmb{{q}}_{ev}, l) = 0 $, $ k_1 $, $ k_2>0 $, 令$ \overline{r}_i = {\omega}_{e}+k_1{{q}}_{ev}+k_2|{q}_{ei}|^l {\rm{sgn}}({q}_{ei}) $, $ {\pmb{f}} $函数定义如式(14) (见本页下方). 式(14)中, $ l = $ $ g/h $, $ \overline{\mu} $为很小的正数, $ \zeta_1 = (2-l)\overline{\mu}^{l-1} $, $ \zeta_2 = (l $ $ - $ $ 1)\overline{\mu}^{l-2} $, $ g $和$ h $均为正奇数且满足$ l\in(0, 1) $.

    假设1[22-23]. 定义$ \pmb{{d}} = {\pmb{T}}_d-{{C}}^{\rm T}\ddot{\boldsymbol{\eta}} $为总干扰, 则$ \pmb{{d}} $满足$ \|\pmb{{d}}\|\leq b_1+b_2\|\boldsymbol{\omega}\|^2 $且$ b_1 $, $ b_2>0 $, $ \|\cdot\| $表示向量2范数. 因此可得不等式$ \|\pmb{{d}}\|\leq p_m\phi $, 这里$ \phi $ $ = $ $ 1+\|\boldsymbol{\omega}\|^2 $且未知$ p_m = \max(b_1, b_2) $.

    本文控制目标为: 针对式(1)~(5)描述的以SGCMGs为执行机构的挠性航天器系统, 当存在执行机构扰动力矩、参数不确定性及外部干扰时, 在假设1的条件下, 设计有限时间自适应鲁棒控制器, 使得挠性航天器闭环控制系统姿态角和姿态角速度误差在有限时间内收敛到原点附近的邻域内, 即

    $$ \begin{align*} \lim\limits_{t \to T}\|\pmb{{q}}_{ev}(t)\|\leq{\varGamma_a}, \ \ \lim\limits_{t \to T}\|\boldsymbol{\omega}_e(t)\|\leq{\varGamma_b} \end{align*} $$

    其中, $ T $为收敛时间, $ \varGamma_a $, $ \varGamma_b>0 $.

    对式(13)进行求导可得

    $$ \begin{align} \dot{\pmb{{r}}} = \dot{\boldsymbol{\omega}}_{e}+k_1\dot{\pmb{{q}}}_{ev}+k_2\dot{{\pmb{f}}}(\pmb{{q}}_{ev}, l) \end{align} $$ (15)

    其中

    $$ \begin{align*} &\dot{{\pmb{f}}}(\pmb{{q}}_{ev}, l) = \\& \begin{cases} l {\rm{diag}}\{|{{q}}_{e1}|^{l-1}, \; \; |{{q}}_{e2}|^{l-1}, \; \; |{{q}}_{e3}|^{l-1}\} {\pmb{{\dot q}}}_{ev}, \\ \qquad \qquad \qquad\qquad\qquad\qquad\quad\ {\overline{r}_i} = 0\; \mbox{或}\; {\overline{r}_i}\neq 0, \\ \qquad \qquad \qquad\qquad\qquad\qquad\quad\ \|{q}_{ei}\|\geq\overline{\mu}\\[-1mm] \zeta_1\dot{\pmb{{q}}}_{ev}+2\zeta_2\, \times\\ {\rm{diag}} \{{q}_{e1} {\rm{sgn}}({q}_{e1}), \ {q}_{e2} {\rm{sgn}}({q}_{e2}), \ {q}_{e3} {\rm{sgn}} ({q}_{e3})\}{\pmb{{\dot q}}}_{ev}, \\\qquad \qquad \qquad\qquad\qquad\qquad\quad\ {\overline{r}_i}\neq 0, \|{q}_{ei}\|<\overline{\mu} \end{cases} \end{align*} $$
    $$ \begin{align*} {\pmb{f}}(\pmb{{q}}_{ev}, l) = \begin{cases} [|{q}_{e1}|^l {\rm{sgn}}({q}_{e1})\; |{q}_{e2}|^l {\rm{sgn}} ({q}_{e2})\; |{q}_{e3}|^l {\rm{sgn}}({q}_{e3})]^{{\rm T}}, &{\overline{r}_i} = 0\; \mbox{或}\; {\overline{r}_i}\neq 0, \|{q}_{ei}\|\geq\overline{\mu}\; \; \; \; \; \; \, {\rm(14a)} \\ \zeta_1\pmb{{q}}_{ev}+\zeta_2 {\rm{diag}} \{{q}_{e1} {\rm{sgn}}({q}_{e1}), {q}_{e2} {\rm{sgn}}({q}_{e2}), {q}_{e3} {\rm{sgn}}({q}_{e3})\}\pmb{{q}}_{ev}, &{\overline{r}_i}\neq 0, \|{q}_{ei}\|<\overline{\mu}\; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \, { \rm{(14b)}} \end{cases} \end{align*} $$

    对式(15) 两边左乘$ {{J}} $可得

    $$ \begin{align} {{J}}\dot{\pmb{{r}}} = {{J}}\dot{\boldsymbol\omega}-{{J}}\dot{\boldsymbol{\omega}}_{r}+ k_1{{J}}\dot{\pmb{{q}}}_{ev}+k_2{{J}}\dot{{\pmb{f}}}(\pmb{{q}}_{ev}, l) \end{align} $$ (16)

    将式(4)代入上式可得

    $$ \begin{align} {{J}}\dot{\pmb{{r}}} = &-\omega^{\times}{{J}}\boldsymbol\omega-{{J}}\dot{\boldsymbol{\omega}}_{r}+ k_1{{J}}\dot{\pmb{{q}}}_{ev}\; +\\ &\ \pmb{{d}}+{\pmb{T}}_{\rm cmg}+k_2{{J}}\dot{{\pmb{f}}}(\pmb{{q}}_{ev}, l) \end{align} $$ (17)

    由于$ {{J}} $中可能存在不确定性, 定义$ \hat{\boldsymbol{\theta}}_J $为$ {\boldsymbol{\theta}}_J = [J_{11} $ $ J_{22} $ $ J_{33} $ $ J_{23} $ $ J_{13} $ $ J_{12}] $的估计值. 定义$ {{L}} $: $ \textbf{R}^{{3}\times{1}}\rightarrow $ $ \textbf{R}^{{3}\times \rm{6}} $满足

    $$ \begin{align} \pmb{{L}}({{A}}) = \left[ {\begin{array}{*{20}{c}} {a_{1}}&{0}&{0}&{0}&{a_{3}}&{a_{2}}\\ {0}&{a_{2}}&{0}&{a_{3}}&{0}&{a_{1}}\\ {0}&{0}&{a_{3}}&{a_{2}}&{a_{1}}&{0} \end{array}} \right] \end{align} $$ (18)

    其中, $ {{A}} = [a_{1}\ \, a_{2}\ \, a_{3}]^{{\rm T}} $, 因此有$ {J}\pmb{{a}} = {{L}}({{A}})\boldsymbol{\theta}_J $. 定义

    $$ \begin{align*} {{F}}_1 = &-{{L}}(\dot{\boldsymbol{\omega}}_{r})-\omega^{\times}{{L}}(\boldsymbol\omega) \; +\\ &\ {{L}}(k_1\dot{\pmb{{q}}}_{ev})+{{L}}(k_2{{J}}\dot{{\pmb{f}}}(\pmb{{q}}_{ev}, l)) \end{align*} $$

    则由式(11), 闭环控制系统方程(17)可写为

    $$ \begin{align} {{J}}\dot{\pmb{{r}}} = &\ {{F}}_1\boldsymbol{\theta}_J+\pmb{{d}}-(h_a{{A}}+{A}{\pmb{F}}_\sigma)\dot{\boldsymbol\delta}\; -\\ &\ {\omega}^{\times}\boldsymbol{h}_{\rm cmg}-{A}{\pmb{f}}+{A}{\pmb{F}}_d \end{align} $$ (19)

    定义摩擦参数不确定性$ \theta_\sigma = \sigma $, $ \hat{{\theta}}_\sigma $为其估计值, 定义估计误差$ \tilde{{\theta}}_{\sigma} = {\theta}_{\sigma}-\hat{{\theta}}_{\sigma} $. 未知框架闭环增益$ \theta_{K_{G}} $ $ = K_{G} $, $ \hat{\theta}_{K_{G}} $为其估计值, $ \tilde{{\theta}}_{K_{G}} = {\theta}_{K_{G}}-\hat{{\theta}}_{K_{G}} $. $ {\pmb{F}}_2 $ $ = $ $ {{A}}\dot{\boldsymbol\delta} $且$ {\varPsi}_2 = {{A}}\hat{\boldsymbol{\theta}}_\sigma $, 则有$ {{A}}F_{\sigma}\dot{\boldsymbol{\delta}} = {\pmb{F}}_2{\theta}_\sigma $且$ {\varPsi}_2\dot{\boldsymbol{\delta}} $ $ = $ $ {\pmb{F}}_2\hat{{\theta}}_\sigma $. 同理, $ {\pmb{F}}_3 = {{A}}{{E}}_d\dot{\boldsymbol\delta} $且$ {\varPsi}_3 = \hat{{\theta}}_{K_{G}}{{A}}{{E}}_d $, 则有$ {{A}}{\theta}_{K_{G}}{{E}}_d\dot{\boldsymbol{\delta}} = {\pmb{F}}_3{{\theta}}_{K_{G}} $且$ {\varPsi}_3\dot{\boldsymbol{\delta}} = {\pmb{F}}_3\hat{{\theta}}_{K_{G}} $. 令$ {B} = {\varPsi}_2 $ $ - $ $ {\varPsi}_3+h_a{{A}} $, 可得

    $$ \begin{align} \label{20} {{J}}\dot{\pmb{{r}}} = &\ {{F}}_1\boldsymbol{\theta}_J- {\pmb{F}}_2{\theta}_\sigma +{{A}}\hat{{\theta}}_\sigma\dot{\boldsymbol{\delta}}-{{A}} \hat{{\theta}}_\sigma\dot{\boldsymbol{\delta}} \, +\\&\ {\pmb{F}}_3{{\theta}}_{K_{G}}+ {{A}}\hat{{\theta}}_{K_{G}}{{E}}_d\dot{\boldsymbol{\delta}} -{{A}}\hat{{\theta}}_{K_{G}}{{E}}_d\dot{\boldsymbol{\delta}} \, -\\&\ h_a{{A}}\dot{\boldsymbol{\delta}}- {A}{\pmb{f}}+ {\pmb{d}} -{\omega}^{\times}\boldsymbol{h}_{\rm cmg} = \\&\ {{F}}_1 \boldsymbol{\theta}_J-{\pmb{F}}_2{\theta}_\sigma+ {\pmb{F}}_2\hat{{\theta}}_\sigma+{\pmb{F}}_3 {{\theta}}_{K_{G}}\, -\\&\ {\pmb{F}}_3\hat{{\theta}}_{K_{G}}-{\varPsi}_2 \dot{\boldsymbol{\delta}}+{\varPsi}_3\dot{\boldsymbol{\delta}} -h_a{{A}}\dot{\boldsymbol{\delta}}-{A}{\pmb{f}}\, + \\&\ {\pmb{d}} -{\omega}^{\times}\boldsymbol{h}_{\rm cmg} = {{F}}_1\boldsymbol{\theta}_J- {\pmb{F}}_2{\tilde{\theta}}_\sigma +{\pmb{F}}_3\tilde{\theta}_{K_{G}}\, - \end{align} $$
    $$ \begin{align} &\ ({\varPsi}_2-{\varPsi}_3+h_a{{A}})\dot{\boldsymbol{\delta}}-{A}{\pmb{f}}\, + \\&\ {\pmb{d}} -{\omega}^{\times}\boldsymbol{h}_{\rm cmg} = {{F}}_1\boldsymbol{\theta}_J-{\pmb{F}}_2{\tilde{\theta}}_\sigma +{\pmb{F}}_3\tilde{\theta}_{K_{G}}\, -\\&\ {B}\dot{\boldsymbol{\delta}}-{A}{\pmb{f}}+ {\pmb{d}} -{\omega}^{\times}\boldsymbol{h}_{\rm cmg} \end{align} $$ (20)

    设计基于终端滑模的有限时间自适应鲁棒控制律

    $$ \begin{align} { \pmb{u}} = \pmb{{u}}_c+\pmb{{u}}_r \end{align} $$ (21)

    其中, $ \pmb{{u}}_c $为自适应模型补偿项, $ \pmb{{u}}_r $为趋近控制项.

    设计趋近控制项为

    $$ \begin{align} &\pmb{{u}}_r = -k_3\pmb{{r}}-k_4{\pmb{f}}_1(\pmb{{r}}, l_1)\\ &{\pmb{f}}_1(\pmb{{r}}, l_1) = \left[|{r}_1|^{l_1}{\rm sgn}({r}_1)\ \, |{r}_2|^{l_1}{\rm sgn}({r}_2)\right.\\&\qquad\qquad\quad \left.|{r}_3|^{l_1}{\rm sgn}({r}_3)\right]^{{\rm T}} \end{align} $$ (22)

    其中, $ k_3 $, $ k_4>0 $, $ l_1 = g_1/h_1 $, $ g_1 $和$ h_1 $均为正奇数且满足$ l_1\in(0, 1) $.

    令$ \hat{\theta}_1 $和$ \hat{\theta}_2 $分别为$ p_m $, $ f_m $估计值, 设计式(21)中的模型补偿项$ \pmb{{u}}_c $为

    $$ \begin{align} \pmb{{u}}_c = -{{F}}_1\hat{\boldsymbol{\theta}}_J+{\omega}^{\times}\boldsymbol{h}_{\rm cmg} -\frac{\phi\hat{\theta}_1\pmb{{r}}}{\|\pmb{{r}}\|+\sigma_1} -\frac{2A_m\hat{\theta}_2\pmb{{r}}}{\|\pmb{{r}}\|+\sigma_2} \end{align} $$ (23)

    模型补偿项$ \pmb{{u}}_c $用于补偿系统惯量参数不确定性、SGCMGs摩擦力矩及总外部干扰. 其中

    $$ \begin{align} \sigma_1 = \frac{\mu_a}{1+\phi}, \quad \sigma_2 = \frac{\mu_b}{1+2A_m} \end{align} $$ (24)

    这里$ \mu_a $, $ \mu_b $取很小的正数, 与系统稳态误差相关.

    在不需要参数先验信息的情况下, 参数更新律分别设计为

    $$ \begin{align} \begin{cases} \dot{\hat{\boldsymbol{\theta}}}_J = {\varGamma}_J{{F}}_1^{{\rm T}}\pmb{{r}} -{\varGamma}_J\gamma_J{\hat{\boldsymbol{\theta}}}_J\\ \dot{\hat{{\theta}}}_\sigma = -{\varGamma}_\sigma{\pmb{F}}_2^{{\rm T}}\pmb{{r}} -{\varGamma}_\sigma\gamma_\sigma{\hat{{\theta}}}_\sigma\\ \dot{\hat{{\theta}}}_{K_{G}} = {\varGamma}_{K_{G}}{\pmb{F}}_3^{{\rm T}}\pmb{{r}} -{\varGamma}_{K_{G}}\gamma_{K_{G}}{\hat{{\theta}}}_{K_{G}}\\ \dot{\hat{\theta}}_1 = \dfrac{\varGamma_1\phi\|\pmb{{r}}\|^2}{\|\pmb{{r}}\|+\sigma_1}-\varGamma_1\gamma_1\hat{\theta}_1\\ \dot{\hat{\theta}}_2 = \dfrac{2\varGamma_2A_m\|\pmb{{r}}\|^2}{\|\pmb{{r}}\|+\sigma_2}-\varGamma_2\gamma_2\hat{\theta}_2 \end{cases} \end{align} $$ (25)

    其中, $ \varGamma_J $, $ \varGamma_\sigma $, $ \varGamma_{K_{G}} $, $ \varGamma_i $, $ \gamma_i $ $ (i = 1, 2) $, $ \gamma_J $, $ \gamma_\sigma $, $ \gamma_{K_{G}} $均为正系数.

    本文金字塔构型SGCMGs采用加权伪逆操纵律[19]

    $$ \begin{align} \dot{\boldsymbol{\delta}} = -{{WB}}^{{\rm T}}[{{BWB}}^{{\rm T}}]^{-1}\pmb{{u}} \end{align} $$ (26)

    其中

    $$ \begin{align*} \ {W} = \alpha \left[ {\begin{array}{*{20}{c}} {W_1}&{\alpha}&{\alpha}&{\alpha}\\ {\alpha}&{W_2}&{\alpha}&{\alpha}\\ {\alpha}&{\alpha}&{W_3}&{\alpha}\\ {\alpha}&{\alpha}&{\alpha}&{W_4} \end{array}} \right] \end{align*} $$

    $ W_i>0 $ $ (i = 1, 2, 3, 4) $, $ \alpha = \alpha_0\exp^{-k_0D} $. $ D = \det({{BB}}^{{\rm T}}) $为奇异度量. $ k_0 $, $ \alpha_0> \rm{0} $.

    注2. 式(26)中的等效雅可比矩阵$ B = {\varPsi}_2 $ $ - $ $ {\varPsi}_3+h_a{{A}} $, 通过对摩擦参数不确定性$ \theta_\sigma $及未知框架闭环增益$ \theta_{K_{G}} $的估计, 来补偿SGCMGs部分摩擦力矩及电磁干扰力矩.

    基于SGCMGs的挠性航天器有限时间自适应鲁棒控制器结构图如图 2所示.

    图 2  基于SGCMGs的挠性航天器有限时间自适应鲁棒控制器结构图
    Fig. 2  Structure diagram of flnite-time adaptive robust control for SGCMGs-based flexible spacecraft

    在上文基础上, 可得以下定理:

    定理1. 针对式(4)~(6)描述的以SGCMGs为执行机构的挠性航天器系统, 当存在执行机构扰动力矩$ {\pmb{F}}_f $和$ {\pmb{F}}_d $、参数不确定性$ {{J}} $及外部干扰$ {\pmb{T}}_{d} $时, 在假设1的条件下, 采用有限时间自适应鲁棒控制器(21)~(24)、参数更新律(25)和操纵律(26), 挠性航天器姿态闭环控制系统实际有限时间稳定, 即姿态角误差和姿态角速度误差在有限时间内收敛到原点附近的任意小的邻域内, 即

    $$ \begin{align*} \lim\limits_{t \to T}\|\pmb{{q}}_{ev}(t)\|\leq{\varGamma_a}, \ \ \lim\limits_{t \to T}\|\boldsymbol{\omega}_e(t)\|\leq{\varGamma_b} \end{align*} $$

    其中, $ T $为收敛时间, $ \varGamma_a $, $ \varGamma_b>0 $.

    证明.

    1) 证明系统到达滑模面后, 姿态角误差可以在有限时间内收敛到零. 选择候选Lyapunov函数为

    $$ \begin{align*} V_0 = \dfrac{1}{2}\left[q_{e1}^2+q_{e2}^2+q_{e3}^2+(1-q_{e4})^2\right] \end{align*} $$

    由于$ q_{e1}^2+q_{e2}^2+q_{e3}^2+q_{e4}^2 = 1 $, $ V_0 = \dfrac{1}{2}[1-q_{e4}^2 $ $ + $ $ (1 $ $ - $ $ q_{e4})^2] $, 由系统运动学方程, $ V_0 $对时间求导可得

    $$ \begin{align*} \dot{V}_0 = \dfrac{1}{2}\pmb{{q}}_{ev}^{{\rm T}}\boldsymbol{\omega}_{e} \end{align*} $$

    系统到达滑模面后$ \pmb{{r}} = 0 $, 满足式(14a), 此时有$ \boldsymbol{\omega}_{e} $ $ = $ $ -k_1\pmb{{q}}_{ev}-k_2{\pmb{f}}(\pmb{{q}}_{ev}, l) $, 代入上式可得

    $$ \begin{align*} \dot{V}_0 = -\frac{k_1}{2}\pmb{{q}}_{ev}^{{\rm T}}\pmb{{q}}_{ev}-\frac{k_2}{2}\pmb{{q}}_{ev}^{{\rm T}}{\pmb{f}} (\pmb{{q}}_{ev}, l)\leq0 \end{align*} $$

    当$ q_{ei}\neq0 $时, $ \dot{V}_0<0 $, 由$ {V}_0 $可得系统状态渐近稳定且$ \pmb{{q}}_{ev} = \textbf{0} $, $ q_{e4} = 1 $是系统平衡点. 经过时间$ T_a $, 存在$ (1-q_{e4})^2\leq(1-q_{e4})(1+q_{e4}) = q_{e1}^2+q_{e2}^2+q_{e3}^2 $, 因此有$ {V}_0\leq q_{e1}^2+q_{e2}^2+q_{e3}^2 $. 由文献[24]中引理不等式, 对任意实数$ x_i\in {\bf R} $, $ i = 1, 2, \cdots, n $, $ p\in(0, 1) $, 有$ (|x_1|+\cdots+|x_n|)^p\leq|x_1|^p+\cdots+|x_n|^p $, 可得

    $$ \begin{align*} &\pmb{{q}}_{ev}^{{\rm T}}{\pmb{f}}(\pmb{{q}}_{ev}, l) = |q_{e1}|^{l+1}+|q_{e2}|^{l+1}+|q_{e3}|^{l+1} = \\ &\qquad |q_{e1}|^{2\frac{1+1}{2}}+|q_{e2}|^{2\frac{1+1}{2}}+|q_{e3}|^{2\frac{1+1}{2}}\geq\\ &\qquad (q_{e1}^2+q_{e2}^2+q_{e3}^2)^{\frac{1+1}{2}}\geq{V}_0^{\frac{1+1}{2}} \end{align*} $$

    因此有$ \dot{V}_0+\frac{k_1}{2}\dot{V}_0+\frac{k_2}{2}\dot{V}_0^{\frac{1+1}{2}}\leq0 $. 由文献[25]中引理可得, $ \lim_{t \to T_b}\pmb{{q}}_{ev} = 0 $和$ \lim_{t \to T_b}{{q}}_{e4} = 1 $可以在有限时间$ T_b $之内实现, 且$ T_b $满足

    $$ \begin{align*} T_b\leq \dfrac{4}{k_1(1-l)}\ln\dfrac{k_1|(V_0|_{t = 0})|^{1-\frac{l+1}{2}}}{k_2} \end{align*} $$

    其中, $ V_0|_{t = 0} $为$ V_0 $在$ t = 0 $时刻的值. 根据终端滑模面定义式(13)可知, 当系统状态到达滑模面后, 在有限时间$ T_0 = T_a+T_b $, 一旦$ \lim_{t \to T_b}\pmb{{q}}_{ev} = \textbf{0} $, 则有$ \lim_{t \to T_b}\boldsymbol{\omega}_{e} $ $ = \textbf{0} $.

    2) 证明系统状态能够在有限时间内到达收敛至滑模面附近邻域.

    设计候选Lyapunov函数

    $$ \begin{align*} V = \frac{1}{2}\pmb{{r}}^{{\rm T}}{{J}}\pmb{{r}} +\frac{1}{2}\tilde{\boldsymbol{\theta}}^{{\rm T}}{\varGamma}^{-1}\tilde{\boldsymbol{\theta}} \end{align*} $$

    其中

    $$ \begin{align*} &\tilde{\boldsymbol{\theta}}^{{\rm T}} = [\tilde{\boldsymbol{\theta}}_J^{{\rm T}}\ \, \tilde{\theta}_\sigma\ \, \tilde{\theta}_{K_G}\ \, \tilde{\theta}_1\ \, \tilde{\theta}_2]\\ &\varGamma = \textbf{I}_{10}[{\varGamma}_J\ \, {\varGamma}_J\ \, {\varGamma}_J\ \, {\varGamma}_J\ \, {\varGamma}_J\ \, {\varGamma}_J\ \, {\varGamma}_\sigma\ \, {\varGamma}_{K_G}\ \, {\varGamma}_1\ \, {\varGamma}_2] \end{align*} $$

    定义参数估计误差分别为$ \tilde{\boldsymbol{\theta}}_J = \boldsymbol{\theta}_J-\hat{\boldsymbol{\theta}}_J $, $ \tilde{{\theta}}_\sigma = {\theta}_\sigma- $ $ \hat{{\theta}}_\sigma $, $ \tilde{{\theta}}_{K_G} = {\theta}_{K_G}-\hat{{\theta}}_{K_G} $, $ \tilde{\theta}_i = \theta_i-\hat{\theta}_i $ $ (i = 1, 2) $, 则由定义有$ \dot{\tilde{\boldsymbol{\theta}}}_J = -\dot{\hat{\boldsymbol{\theta}}}_J $, $ \dot{\tilde{{\theta}}}_\sigma = -\dot{\hat{{\theta}}}_\sigma $, $ \dot{\tilde{{\theta}}}_{K_G} = -\dot{\hat{{\theta}}}_{K_G} $, $ \dot{\tilde{\theta}}_i $ $ = $ $ -\dot{\hat{\theta}}_i $ $ (i = 1, 2) $.

    沿系统(20)轨迹, 对Lyapunov函数求导, 并将式(21)$ \, \sim\, $(26)代入得

    $$ \begin{align*} \dot{V} = &\ \pmb{{r}}^{\rm T}{{J}}\dot{\pmb{{r}}} -\tilde{\boldsymbol{\theta}}_J^{{\rm T}}{\varGamma}_J^{-1} \dot{\hat{\boldsymbol{\theta}}}_J -\tilde{{\theta}}_\sigma^{{\rm T}}{\varGamma}_\sigma^{-1} \dot{\hat{{\theta}}}_\sigma \, -\\ &\ \tilde{{\theta}}_{K_G}^{{\rm T}}{\varGamma}_{K_G}^{-1} \dot{\hat{{\theta}}}_{K_G} -\varGamma_1^{-1}\tilde{\theta}_1\dot{\hat{\theta}}_1 -\varGamma_2^{-1}\tilde{\theta}_2\dot{\hat{\theta}}_2 = \\ &\ \pmb{{r}}^{{\rm T}}\Big({{F}}_1\boldsymbol{\theta}_J -{\pmb{F}}_2\tilde{{\theta}}_\sigma+{\pmb{F}}_3 \tilde{{\theta}}_{K_G}+\pmb{{u}} -{A}{\pmb{f}}\, +\\ &\ {\pmb{{d}}}- {\omega}^{\times}\boldsymbol{h}_{\rm cmg}\Big)-\tilde{\boldsymbol{\theta}}_J^{{\rm T}}{\varGamma}_J^{-1} \dot{\hat{\boldsymbol{\theta}}}_J -\tilde{{\theta}}_\sigma^{{\rm T}}{\varGamma}_\sigma^{-1} \dot{\hat{{\theta}}}_\sigma \, -\\ &\ \tilde{{\theta}}_{K_G}^{{\rm T}}{\varGamma}_{K_G}^{-1} \dot{\hat{{\theta}}}_{K_G} -\varGamma_1^{-1}\tilde{\theta}_1\dot{\hat{\theta}}_1 -\varGamma_2^{-1}\tilde{\theta}_2\dot{\hat{\theta}}_2 = \\ &\ \pmb{{r}}^{{\rm T}}\Big(-k_3\pmb{{r}}- k_4{\pmb{f}}_1(\pmb{{r}}, l_1) +{{F}}_1\tilde{\boldsymbol{\theta}}_J-{\pmb{F}}_2 \tilde{{\theta}}_\sigma\, +\\ &\ {\pmb{F}}_3\tilde{{\theta}}_{K_G} -\frac{\phi\hat{\theta}_1\pmb{{r}}}{\|\pmb{{r}}\|+\sigma_1} -\frac{2A_m\hat{\theta}_2\pmb{{r}}}{\|\pmb{{r}}\|+\sigma_2} -{A}{\pmb{f}}+{\pmb{{d}}} \Big)\, -\\ &\ \tilde{\boldsymbol{\theta}}_J^{{\rm T}}{{F}}_1^{{\rm T}} \pmb{{r}}+ \tilde{\boldsymbol{\theta}}_J^{{\rm T}}\gamma_J \hat{\boldsymbol{\theta}}_J+ \tilde{{\theta}}_\sigma^{{\rm T}}{\pmb{F}}_2^{{\rm T}} \pmb{{r}}+ \tilde{{\theta}}_\sigma^{{\rm T}}\gamma_\sigma \hat{{\theta}}_\sigma\, - \end{align*} $$
    $$ \begin{align*} & \tilde{{\theta}}_{K_G}^{{\rm T}}{\pmb{F}}_3^{{\rm T}} \pmb{{r}}+ \tilde{{\theta}}_{K_G}^{{\rm T}}\gamma_{K_G} \hat{{\theta}}_{K_G}-\tilde{\theta}_1\frac{\phi\|\pmb{{r}}\|^2} {\|\pmb{{r}}\|+ \sigma_1}\, +\\ &\gamma_1\tilde{\theta}_1\hat{\theta}_1- \tilde{\theta}_2\frac{2A_m\|\pmb{{r}}\|^2} {\|\pmb{{r}}\|+ \sigma_2}+\gamma_2\tilde{\theta}_2\hat{\theta}_2 = \\ &\pmb{{r}}^{{\rm T}}{\pmb{{d}}}- k_3\pmb{{r}}^{{\rm T}}\pmb{{r}} -k_4\pmb{{r}}^{{\rm T}}{\pmb{f}}_1 (\pmb{{r}}, l_1) -\pmb{{r}}^{{\rm T}}{A}{\pmb{f}} \, -\\&\frac{\phi\hat{\theta}_1\|\pmb{{r}}\|^2} {\|\pmb{{r}}\|+ \sigma_1} -\frac{\phi\tilde{\theta}_1\|\pmb{{r}}\|^2} {\|\pmb{{r}}\|+ \sigma_1} -\frac{2A_m\hat{\theta}_2\|\pmb{{r}}\|^2} {\|\pmb{{r}}\|+ \sigma_2} \, -\\ &\frac{2A_m\tilde{\theta}_2\|\pmb{{r}}\|^2} {\|\pmb{{r}}\|+ \sigma_2}+\gamma_J\tilde{\boldsymbol{\theta}}_J^{{\rm T}} \hat{\boldsymbol{\theta}}_J+ \gamma_\sigma\tilde{{\theta}}_\sigma^{{\rm T}} \hat{{\theta}}_\sigma\, +\\ & \gamma_{K_G}\tilde{{\theta}}_{K_G}^{{\rm T}} \hat{{\theta}}_{K_G}+ \gamma_1\tilde{\theta}_1\hat{\theta}_1+ \gamma_2\tilde{\theta}_2\hat{\theta}_2 \end{align*} $$

    由文献[24]中引理不等式可得, $ \pmb{{r}}^{{\rm T}}{\pmb{f}}_1(\pmb{{r}}, l_1)\geq\|\pmb{{r}}\|^{l_1+1} $, 因此有

    $$ \begin{align} \dot{V}\leq &-k_4\|\pmb{{r}}\|^{l_1+1}+\phi{\theta}_1\| \pmb{{r}}\|+2A_m{\theta}_2\|\pmb{{r}}\|\, -\\ &\ \frac{\phi\hat{\theta}_1\|\pmb{{r}}\|^2}{\|\pmb{{r}}\| +\sigma_1} -\frac{\phi\tilde{\theta}_1\|\pmb{{r}}\|^2}{\|\pmb{{r}}\| +\sigma_1} -\frac{2A_m\hat{\theta}_2\|\pmb{{r}}\|^2}{\|\pmb{{r}}\| +\sigma_2}\, -\\ &\ \frac{2A_m\tilde{\theta}_2\|\pmb{{r}}\|^2}{\|\pmb{{r}}\| +\sigma_2}+ \gamma_J\tilde{\boldsymbol{\theta}}_J^{{\rm T}}\hat{\boldsymbol{\theta}}_J+ \gamma_\sigma\tilde{{\theta}}_\sigma^{{\rm T}}\hat{{\theta}}_\sigma\, +\\ &\ \gamma_{K_G}\tilde{{\theta}}_{K_G}^{{\rm T}}\hat{{\theta}}_{K_G}+ \gamma_1\tilde{\theta}_1\hat{\theta}_1+ \gamma_2\tilde{\theta}_2\hat{\theta}_2 \end{align} $$ (27)

    令$ w = \phi{\theta}_1\|\pmb{{r}}\|+2A_m{\theta}_2\|\pmb{{r}}\|- \frac{\phi\hat{\theta}_1\|\pmb{{r}}\|^2}{\|\pmb{{r}}\|+\sigma_1} -\frac{\phi\tilde{\theta}_1\|\pmb{{r}}\|^2}{\|\pmb{{r}}\|+\sigma_1} -\frac{2A_m\hat{\theta}_2\|\pmb{{r}}\|^2}{\|\pmb{{r}}\|+\sigma_2}- \frac{2A_m\tilde{\theta}_2\|\pmb{{r}}\|^2}{\|\pmb{{r}}\|+\sigma_2} $, 又由式(24), 可得

    $$ \begin{align} w\leq &\ \frac{\phi\theta_1\|\pmb{{r}}\|\sigma_1+ \phi\theta_1\sigma_1^2}{\|\pmb{{r}}\|+\sigma_1} \, +\\ &\ 2\frac{A_m\theta_2\|\pmb{{r}}\|\sigma_2+ A_m\theta_2\sigma_2^2}{\|\pmb{{r}}\|+\sigma_2} = \\ &\ \frac{\phi}{1+\phi}\mu_a\theta_1+\frac{2A_m}{1+2A_m} \mu_b\theta_2\leq\mu_a\theta_1+\mu_b\theta_2 \end{align} $$ (28)

    由文献[26]中, 航天器正定对称惯量矩阵$ {{J}} $满足不等式$ \lambda_{J\min}\|\pmb{{x}}\|^2\leq \pmb{{x}}^{{\rm T}}{{J}}\pmb{{x}}\leq\lambda_{J\max}\|\pmb{{x}}\|^2 $, 对$ \forall\pmb{{x}} $ $ \in $ $ \textbf{R}^{3\times1} $, $ \lambda_{J\max} $和$ \lambda_{J\min} $为正常数. 令$ p_1 = (l_1+ $ $ 1)/{2} $, 对式(27)进一步放缩可得

    $$ \begin{align*} \dot{V}\leq & -k_4\|\pmb{{r}}\|^{2p_1}+w+ \gamma_J\tilde{\boldsymbol{\theta}}_J^{{\rm T}}\hat{\boldsymbol{\theta}}_J+ \gamma_\sigma\tilde{{\theta}}_\sigma\hat{{\theta}}_\sigma\, +\\ &\ \gamma_{K_G}\tilde{{\theta}}_{K_G}\hat{{\theta}}_{K_G}+ \gamma_1\tilde{\theta}_1\hat{\theta}_1+\gamma_2\tilde{\theta}_2 \hat{\theta}_2\leq\\ &-\dfrac{k_42^{p_1}}{\lambda_{J\max}^{p_1}}\left(\dfrac{1}{2} \pmb{{r}}^{{\rm T}}{{J}}\pmb{{r}}\right)^{p_1} -\left(\dfrac{\gamma_1(2\mu_1-1)}{2\mu_1}\tilde{\theta}_1^2\right)^{p_1}\! -\\ &\ \left(\dfrac{\gamma_2(2\mu_2-1)}{2\mu_2}\tilde{\theta}_2^2\right)^{p_1} - \end{align*} $$
    $$ \begin{align*} &\left(\dfrac{\gamma_J(2\mu_J-1)}{2\mu_J}\tilde{\boldsymbol {\theta}}_J^{{\rm T}}\tilde{\boldsymbol{\theta}}_J\right)^{p_1}-\\ &\left(\dfrac{\gamma_\sigma(2\mu_\sigma-1)}{2\mu_\sigma} \tilde{\theta}_\sigma^2\right)^{p_1} -\\ &\left(\dfrac{\gamma_{K_G}(2\mu_{K_G}-1)}{2\mu_{K_G}} \tilde{\theta}_{K_G}^2\right)^{p_1} +\\ &\left(\dfrac{\gamma_1(2\mu_1-1)}{2\mu_1}\tilde{\theta}_1^2\right)^{p_1}+ \gamma_1\tilde{\theta}_1\hat{\theta}_1\, +\\ &\left(\dfrac{\gamma_2(2\mu_2-1)}{2\mu_2}\tilde{\theta}_2^2\right)^{p_1}+ \gamma_2\tilde{\theta}_2\hat{\theta}_2 \, +\\ &\left(\dfrac{\gamma_J(2\mu_J-1)}{2\mu_J}\tilde{\boldsymbol {\theta}}_J^{{\rm T}}\tilde{\boldsymbol{\theta}}_J\right)^{p_1} +\gamma_J\tilde{\boldsymbol{\theta}}_J^{{\rm T}}\hat{\boldsymbol {\theta}}_J \, +\\ &\bigg(\dfrac{\gamma_\sigma(2\mu_\sigma-1)}{2\mu_\sigma}\tilde {\theta}_\sigma^2\bigg)^{p_1} +\gamma_\sigma\tilde{{\theta}}_\sigma\hat{{\theta}}_\sigma\, +\\ &\left(\dfrac{\gamma_{K_G}(2\mu_{K_G}-1)}{2\mu_{K_G}} \tilde{\theta}_{K_G}^2\right)^{p_1} \, +\\ &\gamma_{K_G}\tilde{{\theta}}_{K_G}\hat{{\theta}}_{K_G}+ \mu_a\theta_1+\mu_b\theta_2 \end{align*} $$

    取$ \mu_i>{1}/{2} $, $ i = 1, 2, J, \theta, K_G $. 当$ i = 1, 2, \theta, K_G $, 可得

    $$ \begin{align} \gamma_i\tilde{\theta}_i\hat{\theta}_i = &\ \gamma_i{\theta}_i \tilde{\theta}_i-\gamma_i\tilde{\theta}_i^2\leq\\ &\ \gamma_i\left(-\tilde{\theta}_i^2+\dfrac{1}{2\mu_i} \tilde{\theta}_i^2+\dfrac{\mu_i}{2}{\theta}_i^2\right) = \\ &\ \dfrac{-\gamma_i(2\mu_i-1)}{2\mu_i}\tilde{\theta}_i^2+ \dfrac{\gamma_i\mu_i}{2}\theta_i^2 \end{align} $$ (29)

    令$ m_i = \dfrac{\gamma_i(2\mu_i-1)}{2\mu_i}\tilde{\theta}_i^2 $, 当$ m_i>1 $时, 由式(29)可得

    $$ \begin{align} m_i^{p_1}+\gamma_i\tilde{\theta}_i\hat{\theta}_i\leq m_i-m_i+\dfrac{\gamma_i\mu_i}{2}\theta_i^2 = \dfrac{\gamma_i\mu_i}{2}\theta_i^2 \end{align} $$ (30)

    当$ m_i\leq1 $时, 由式(29)和式(30)可得

    $$ \begin{align*} m_i^{p_1}\big|_{m_i\leq1}<m_i^{p_1}\big|_{m_i>1} \end{align*} $$

    从而可得

    $$ \begin{align} m_i^{p_1}\big|_{m_i\leq1}+\gamma_i\tilde{\theta}_i\hat{\theta}_i<m_i^{p_1}\big|_{m_i>1}+\gamma_i\tilde{\theta}_i\hat{\theta}_i\leq\dfrac{\gamma_i\mu_i}{2}\theta_i^2 \end{align} $$ (31)

    同理, 当$ i = J $时, 令$ m_J = \dfrac{\gamma_J(2\mu_J-1)}{2\mu_J}\tilde{\boldsymbol{\theta}}_J^{{\rm T}}\tilde{\boldsymbol{\theta}}_J $可得

    $$ \begin{align} & \gamma_J\tilde{\boldsymbol{\theta}}_J^{{\rm T}}\hat{\boldsymbol{\theta}}_J\leq \dfrac{-\gamma_J(2\mu_i-1)}{2\mu_i}\tilde{\boldsymbol{\theta}}_J^{{\rm T}}\tilde{\boldsymbol{\theta}}_J +\dfrac{\gamma_i\mu_i}{2}{\boldsymbol{\theta}}_J^{{\rm T}}{\boldsymbol{\theta}}_J \end{align} $$ (32)
    $$ \begin{align} & m_J^{p_1}+\gamma_J\tilde{\boldsymbol{\theta}}_J\hat{\boldsymbol{\theta}}_J \leq\dfrac{\gamma_i\mu_i}{2}{\boldsymbol{\theta}}_J^{{\rm T}}{\boldsymbol{\theta}}_J \end{align} $$ (33)

    令$ \lambda = -{k_42^{p_1}}/{\lambda_{J\max}^{p_1}} $, $ \eta = \sum_{i = 1}^{2}\dfrac{\gamma_i\mu_i}{2}\theta_i^2 $ $ + $ $ \dfrac{\gamma_J\mu_J}{2}{\boldsymbol{\theta}}_J^{{\rm T}}{\boldsymbol{\theta}}_J + \dfrac{\gamma_\sigma\mu_\sigma}{2}\theta_\sigma^2 +\dfrac{\gamma_{K_G}\mu_{K_G}}{2}\theta_{K_G}^2 +\mu_a\theta_1+\mu_b\theta_2 $, $ \varGamma_i $ $ = \dfrac{\mu_i\lambda^{\frac{1}{p_1}}}{\gamma_i(2\mu_i-1)} $, $ i = 1, 2, J, \theta, K_G $, 由式(29)~(33)及文献[24]中引理可得

    $$ \begin{align*} \dot{V}\leq &-\lambda\left(\bigg(\dfrac{\pmb{{r}}^{{\rm T}}{{J}}\pmb{{r}}}{2}\bigg)^{p_1} +\bigg(\dfrac{\tilde{\boldsymbol{\theta}}_J^{{\rm T}}{\varGamma}_J^{-1}\tilde{\boldsymbol{\theta}}_J}{2}\bigg)^{p_1}\, +\right.\\ &\ \bigg(\dfrac{\tilde{\theta}_\sigma^2}{2\varGamma_\sigma}\bigg)^{p_1} +\bigg(\dfrac{\tilde{\theta}_{K_G}^2}{2\varGamma_{K_G}}\bigg)^{p_1}\, +\\ & \left.\bigg(\dfrac{\tilde{\theta}_1^2}{2\varGamma_1}\bigg)^{p_1} +\bigg(\dfrac{\tilde{\theta}_2^2}{2\varGamma_2}\bigg)^{p_1}\right)+\eta\leq\\ &-\lambda\left(\dfrac{\pmb{{r}}^{{\rm T}}{{J}}\pmb{{r}}}{2} +\dfrac{\tilde{\boldsymbol{\theta}}_J^{{\rm T}}{\varGamma}_J^{-1}\tilde{\boldsymbol{\theta}}_J}{2}\right)^{p_1}+\eta = \\ &-\lambda V^{p_1}+\eta \end{align*} $$

    由文献[25]引理可知, 挠性航天器姿态闭环控制系统实际有限时间稳定, 系统轨迹能够在有限时间内到达

    $$ \begin{align*} V^{p_1}\leq\frac{\eta}{(1-\chi)\lambda}, \quad 0<\chi<1 \end{align*} $$

    $ \pmb{{r}} $能够在有限时间内收敛到滑模面附近的邻域内, 即

    $$ \begin{align*} {\lim\limits_{t \to T_1}}\pmb{{r}}(t)\in\left(\|\pmb{{r}}(t)\|\leq \left(\dfrac{\eta}{(1-\chi)k_4}\right)^{\frac{1}{2p_1}} = \varGamma\right) \end{align*} $$

    其中, 收敛时间$ T_1 $满足不等式

    $$ \begin{align*} T_1\leq&\ \frac{(V|_{t = 0})^{1-\frac{1}{2p_1}}}{\left(\lambda\chi\left(1-\dfrac{1}{2p_1}\right)\right)} = \\ &\ \frac{(l_1+1)(V|_{t = 0})^{1-\frac{1}{2p_1}}}{l_1\lambda\chi}, \quad 0<\chi<1 \end{align*} $$

    3) 证明系统姿态角误差$ \pmb{{q}}_{ev} $和姿态角速度误差$ \boldsymbol{\omega}_e $能够在有限时间$ T $内收敛到原点$ (\pmb{{q}}_{ev} = \textbf{0} $, $ \boldsymbol{\omega}_e $ $ = $ $ \textbf{0}) $附近的邻域内. 由式(13)可得当满足式(14a)时

    $$ \begin{align} \begin{cases} r_i = \omega_{ei}+k_1q_{ei}+k_2|q_{ei}|^l = s_i, \ \, |s_i|<\varGamma\\ \omega_{ei}+\Big(k_1-\dfrac{s_i}{2q_{ei}}\Big)q_{ei}+ \Big(k_2-\dfrac{s_i}{2|q_{ei}|^l}\Big)\left|q_{ei}\right|^l = 0\\ \qquad\qquad \qquad\qquad\qquad \qquad\qquad\quad\ \ \, i = 1, 2, 3 \end{cases} \end{align} $$ (34)

    当$ k_1>{s_i}/({2q_{ei}}) $且$ k_2>{s_i}/({2|q_{ei}|^l}) $时, 式(34)与式(13)有相同形式, 结合步骤1)可得, 系统姿态角误差可以在有限时间内收敛到$ |q_{ei}|<\max({s_i}/({2k_1}) $, $ ({s_i}/({2k_2}))^{1/l}) = \varGamma_{ {\rm{m}}i} $, $ i = 1, 2, 3 $. 由式(34)可得, 姿态角速度误差可以在有限时间内收敛到$ |\omega_{ei}| \leq $ $ k_1\varGamma_{{m}i} $ $ + $ $ k_2\varGamma^l_{{m}i}+\varGamma = \varGamma_{{j}i} $.

    当满足式(14b)时, 有

    $$ \begin{align*} r_i = &\ \omega_{ei}+k_1q_{ei}+k_2\zeta_1q_{ei}\, +\\ &\ k_2\zeta_2|q_{ei}|^2 {\rm{sgn}}(q_{ei}) = s_i, \ \ i = 1, 2, 3, \; |s_i|<\varGamma \end{align*} $$

    $ |q_{ei}|\leq\bar{\mu} = \varGamma_{{n}i} $, $ i = 1, 2, 3 $. 由上式得, 角速度误差可在有限时间内收敛到$ |\omega_{ei}|\leq k_1\varGamma_{ {\rm{n}}i}+k_2\varGamma^l_{{n}i}+\varGamma = \varGamma_{{k}i} $.

    综上可得, 姿态角误差和角速度误差在有限时间$ T = T_0+T_1 $内快速收敛到包含原点在内的邻域内, 即

    $$ \begin{align*} \lim\limits_{t \to T}\|\pmb{{q}}_{ev}(t)\|\leq{\varGamma_a}, \; \lim\limits_{t \to T}\|\boldsymbol{\omega}_e(t)\|\leq{\varGamma_b} \end{align*} $$

    其中, $ T $为收敛时间,

    $$ \begin{align*} &\varGamma_a = \max\!\left(\!\sqrt{\varGamma_{m1}^2+\!\varGamma_{m2}^2+\!\varGamma_{m3}^2}, \sqrt{\varGamma_{n1}^2+\!\varGamma_{n2}^2+\!\varGamma_{n3}^2}\right)\\ &\varGamma_b = \max\Big(\sqrt{\varGamma_{k1}^2+\varGamma_{k2}^2+\varGamma_{k3}^2}, \sqrt{\varGamma_{j1}^2+\varGamma_{j2}^2+\varGamma_{j3}^2}\Big) \end{align*} $$

    注3. 可通过合理选择系数$ k_4 $, $ \gamma_i $ $ (i = 1, 2) $, $ \gamma_J $, $ \gamma_\sigma $, $ \gamma_{K_{G}} $, $ \mu_a $和$ \mu_b $, 来减小系统稳态误差, 从而减小系统姿态角误差和姿态角速度误差.

    注4. 本文控制系统中SGCMGs采用加权伪逆操纵律. 为进一步提高系统奇异逃避能力, 在实际应用中可采用广义鲁棒伪逆操纵律[19]

    $$ \begin{align*} &\dot{\boldsymbol{\delta}} = -{{WB}}^{{\rm T}}[{{BWB}}^{{\rm T}} +{{V}}]^{-1}\pmb{{u}}\\ &{{V}} = \alpha \left[ {\begin{array}{*{20}{c}} {1}&{\epsilon}&{\epsilon}\\ {\epsilon}&{1}&{\epsilon}\\ {1}&{\epsilon}&{\epsilon} \end{array}} \right] \end{align*} $$

    其中, $ {B} $, $ {W} $, $ \alpha $同式(26), $ \epsilon = \epsilon_0\sin(\omega_0t) $, $ \epsilon_0 $, $ \omega_0 $ $ > $ $ 0 $. 当SGCMGs接近奇异, 可通过引入力矩误差使系统尽快避开并远离奇异状态; 当SGCMGs系统远离奇异时, 该操纵律可等效为加权伪逆操纵律, 从而精确输出期望力矩. 该操纵律具有更好的奇异逃避能力.

    注5. 本文最终证明闭环控制系统是实际有限时间稳定的, 即稳态误差在有限时间内可收敛至较小的界内, 可根据对航天器姿态机动的精度需求来获得稳定时间. 相较于有限时间稳定及有界稳定, 本文结果对工程应用具有更高的参考价值.

    本节针对式(1)~(6)描述的基于SGCMGs挠性航天器姿态控制系统进行仿真, 挠性航天器的转动惯量$ {{J}} $、挠性附件与本体的耦合矩阵$ {{C}} $等航天器物理参数均取自文献[27]. 外部干扰包含周期性干扰及常值干扰, 形式为

    $$ \begin{align*} &{\pmb{T}}_d = 10^{-4}\, \times \\&\begin{subarray}{l} \begin{bmatrix} -3+4\cos(0.0628t)-\cos(0.1256t)+2\omega_1\sin(0.11t)\\ 4+3\sin(0.0628t)-2\cos(0.1256t)+\omega_2\cos(0.11t)\\ -3+4\sin(0.0628t)-3\sin(0.1256t)-2\omega_3\cos(0.11t) \end{bmatrix} \rm{N}\cdot \rm{m}\end{subarray} \end{align*} $$

    初始姿态角为$ [60^{\circ}\ \, 45^{\circ}\ \, 20^{\circ}]^{{\rm T}} $, 期望姿态角为$ [0^{\circ}\ \, 0^{\circ} $ $ 0^{\circ}]^{{\rm T}} $, 初始及期望姿态角速度均为零. SGCMGs系统倾斜角为$ \beta = 53.13^{\circ} $. 初始框架角为$ \boldsymbol\delta_0 = [0\ \, 0\ \, 0 $ $ 0]^{{\rm T}}\, \rm{rad} $. 初始框架角速度为$ \dot{\boldsymbol\delta_0} = [0\ \, 0\ \, 0\ \, 0]^{{\rm T}}\, \rm{rad/s} $. LuGre摩擦模型参数取自文献[8], 操纵律采用广义鲁棒伪逆操纵律, LuGre摩擦模型参数、控制律及操纵律参数如表 1所示.

    表 1  LuGre摩擦模型参数、控制律及操纵律参数
    Table 1  Parameters of LuGre friction model, control law and steering law
    参数类型 参数值
    LuGre摩擦模型参数及前馈环路增益 $\sigma_1 = 0.3\, \rm{Nm}$, $\sigma_2 = 0.5\, \rm{Nm}$, $\sigma_3 = 0.06\, \rm{Nm}\cdot \rm{s/rad}$,
    $F_ \rm{c} = 0.1\, \rm{Nm}$, $F_ \rm{s} = 0.12\, \rm{Nm}$, $V_ \rm{s} = 0.001\, \rm{rad/s}$, $K_G = 0.2$
    控制律参数 $k_1 = 2.5$, $k_2 = 0.1$, $k_3 = 500$, $k_4 = 0.1$, $g = 11, h = 13, g_1 = 7, h_1 = 9$, $\mu_a = \mu_b = 0.001$,
    $\gamma_J = 0.001, \gamma_{K_G} = \gamma_\sigma = \gamma_1 = \gamma_2 = 2$, $[\varGamma_J\; \varGamma_\sigma\; \varGamma_{K_G}\; \varGamma_1\; \varGamma_2]^{{\rm T}} = [1\; 0.1\; 0.01\; 1\; 0.001]^{{\rm T}}$
    操纵律参数 $W_i = i$ $(i = 1, 2, 3, 4)$, $\omega_0 = 1$, $\epsilon_0 = 10^{-4}$, $k_0 = 10$, $\alpha_0 = 10^{-5}$, $h_a = 100$
    下载: 导出CSV 
    | 显示表格

    对航天器进行初始姿态引导[28]来进一步抑制姿态机动过程中挠性附件的振动. 针对航天器执行机构SGCMGs存在摩擦非线性及电磁扰动力矩, 航天器存在惯量不确定性及外部干扰的情况进行仿真, 仿真结果如图 3~8所示.

    图 3  姿态角响应及稳态放大图
    Fig. 3  Attitude angle and enlarged view of steady state
    图 4  姿态角速度响应及稳态放大图
    Fig. 4  Attitude angular velocity and enlarged view of steady state
    图 5  姿态误差四元数
    Fig. 5  Error of the attitude quaternion
    图 6  挠性附件模态
    Fig. 6  Flexible appendages Modes
    图 7  SGCMGs力矩输出响应
    Fig. 7  Profile of SGCMGs output torque
    图 8  奇异度量
    Fig. 8  Singularity measurement

    由仿真结果图 3~8可得, 当挠性航天器系统存在执行机构非线性干扰力矩、惯量不确定性及外部干扰时, 采用有限时间自适应鲁棒控制器, 可以较好地完成姿态机动控制任务, 航天器三轴姿态角误差和姿态角速度误差可以平滑地在50 s左右分别进入$ \pm5\times10^{-4}\, (^{\circ}) $和$ \pm1\times10^{-4} \, ({^\circ}/ \rm{s}) $误差带, 实现了高精度姿态快速机动控制. 由图 6挠性附件模态可知, 四阶挠性附件振动均得到有效抑制, 并在完成姿态机动后快速衰减到零. 由图 7图 8可得, SGCMGs系统在航天器姿态机动过程中, 虽然有接近奇异的情况, 但系统快速避开了奇异状态, 并在机动结束后远离奇异状态.

    本文设计了一种基于SGCMGs挠性航天器姿态机动有限时间自适应鲁棒控制器, 在减小SGCMGs中摩擦非线性及电磁扰动力矩对系统影响的同时, 针对航天器惯量不确定性进行估计和补偿, 抑制了外部干扰及刚柔耦合效应对系统的影响. 仿真结果表明, 本文所提控制方法可实现挠性航天器高精度快速姿态机动控制.


  • 本文责任编委 倪茂林
  • 图  1  SGCMGs驱动的挠性航天器姿态控制系统结构图

    Fig.  1  Structure diagram of attitude control system for SGCMGs-based flexible spacecraft

    图  2  基于SGCMGs的挠性航天器有限时间自适应鲁棒控制器结构图

    Fig.  2  Structure diagram of flnite-time adaptive robust control for SGCMGs-based flexible spacecraft

    图  3  姿态角响应及稳态放大图

    Fig.  3  Attitude angle and enlarged view of steady state

    图  4  姿态角速度响应及稳态放大图

    Fig.  4  Attitude angular velocity and enlarged view of steady state

    图  5  姿态误差四元数

    Fig.  5  Error of the attitude quaternion

    图  6  挠性附件模态

    Fig.  6  Flexible appendages Modes

    图  7  SGCMGs力矩输出响应

    Fig.  7  Profile of SGCMGs output torque

    图  8  奇异度量

    Fig.  8  Singularity measurement

    表  1  LuGre摩擦模型参数、控制律及操纵律参数

    Table  1  Parameters of LuGre friction model, control law and steering law

    参数类型 参数值
    LuGre摩擦模型参数及前馈环路增益 $\sigma_1 = 0.3\, \rm{Nm}$, $\sigma_2 = 0.5\, \rm{Nm}$, $\sigma_3 = 0.06\, \rm{Nm}\cdot \rm{s/rad}$,
    $F_ \rm{c} = 0.1\, \rm{Nm}$, $F_ \rm{s} = 0.12\, \rm{Nm}$, $V_ \rm{s} = 0.001\, \rm{rad/s}$, $K_G = 0.2$
    控制律参数 $k_1 = 2.5$, $k_2 = 0.1$, $k_3 = 500$, $k_4 = 0.1$, $g = 11, h = 13, g_1 = 7, h_1 = 9$, $\mu_a = \mu_b = 0.001$,
    $\gamma_J = 0.001, \gamma_{K_G} = \gamma_\sigma = \gamma_1 = \gamma_2 = 2$, $[\varGamma_J\; \varGamma_\sigma\; \varGamma_{K_G}\; \varGamma_1\; \varGamma_2]^{{\rm T}} = [1\; 0.1\; 0.01\; 1\; 0.001]^{{\rm T}}$
    操纵律参数 $W_i = i$ $(i = 1, 2, 3, 4)$, $\omega_0 = 1$, $\epsilon_0 = 10^{-4}$, $k_0 = 10$, $\alpha_0 = 10^{-5}$, $h_a = 100$
    下载: 导出CSV
  • [1] Du H B, Li S H. Attitude synchronization for flexible spacecraft with communication delays. IEEE Transactions on Automatic Control, 2016, 61(11): 3625-3630 doi: 10.1109/TAC.2016.2525933
    [2] Zhong C X, Chen Z Y, Guo Y. Attitude control for flexible spacecraft with disturbance rejection. IEEE Transactions on Aerospace and Electronic Systems, 2017, 53(1): 101-110 doi: 10.1109/TAES.2017.2649259
    [3] Zhong C X, Guo Y, Yu Z, Wang L, Chen Q W. Finite-time attitude control for flexible spacecraft with unknown bounded disturbance. Transactions of the Institute of Measurement and Control, 2016, 38(2): 240-249 doi: 10.1177/0142331214566223
    [4] Zhong C X, Wu L P, Guo J, Guo Y, Chen Z Y. Robust adaptive attitude manoeuvre control with finite-time convergence for a flexible spacecraft. Transactions of the Institute of Measurement and Control, 2016, 40(2): 425-435 http://smartsearch.nstl.gov.cn/paper_detail.html?id=69175779e799cf0d6c7cf1ac0934dc26
    [5] 刘宝玉, 金磊, 贾英宏. 谐波齿轮对大型SGCMG框架转速控制的影响分析. 航天控制, 2014, 32(2): 23-28, 34 doi: 10.3969/j.issn.1006-3242.2014.02.005

    Liu Bao-Yu, Jin Lei, Jia Ying-Hong. Influence analysis of harmonic gear to SGCMG gimbal rate control. Aerospace Control, 2014, 32(2): 23-28, 34 doi: 10.3969/j.issn.1006-3242.2014.02.005
    [6] 吴忠, 吴宏鑫. SGCMG框架伺服系统扰动力矩的分析与抑制. 航天控制, 1998, 16(4): 8-16 https://www.cnki.com.cn/Article/CJFDTOTAL-HTKZ199804001.htm

    Wu Zhong, Wu Hong-Xin. Analysis and attenuation of disturbance torque in SGCMG gimbal servo systems. Aerospace Control, 1998, 16(4): 8-16 https://www.cnki.com.cn/Article/CJFDTOTAL-HTKZ199804001.htm
    [7] 鲁明, 张欣, 李耀华. SGCMG框架伺服系统扰动力矩分析与控制. 中国空间科学技术, 2013, 33(1): 15-20 https://www.cnki.com.cn/Article/CJFDTOTAL-ZGKJ201301002.htm

    Lu Ming, Zhang Xin, Li Yao-Hua. Analysis and control for SGCMG gimbal servo system disturbance torque. Chinese Space Science and Technology, 2013, 33(1): 15-20 https://www.cnki.com.cn/Article/CJFDTOTAL-ZGKJ201301002.htm
    [8] 金磊, 徐世杰. SGCMG框架伺服系统动力学建模与低速控制. 中国空间科学技术, 2010, (6): 1-10 https://www.cnki.com.cn/Article/CJFDTOTAL-ZGKJ201006003.htm

    Jin Lei, Xu Shi-Jie. Dynamic modelling and low-speed control for SGCMG gimbal servo system. Chinese Space Science and Technology, 2010, (6): 1-10 https://www.cnki.com.cn/Article/CJFDTOTAL-ZGKJ201006003.htm
    [9] Hill D. Dynamics and control of spacecraft using control moment gyros with friction compensation. Journal of Guidance, Control, and Dynamics, 2016, 39(10): 2406-2418 doi: 10.2514/1.G001585
    [10] MacKunis W, Dupree K, Fitz-Coy N, Dixon W E. Adaptive satellite attitude control in the presence of inertia and CMG gimbal friction uncertainties. The Journal of the Astronautical Sciences, 2008, 56(1): 121-134 doi: 10.1007/BF03256544
    [11] 田林, 徐世杰. 带变速控制力矩陀螺的航天器自适应姿态控制. 中国空间科学技术, 2012, 32(5): 9-16 https://www.cnki.com.cn/Article/CJFDTOTAL-ZGKJ201205004.htm

    Tian Lin, Xu Shi-Jie. Adaptive attitude control of spacecraft with variable speed control moment gyroscopes. Chinese Space Science and Technology, 2012, 32(5): 9-16 https://www.cnki.com.cn/Article/CJFDTOTAL-ZGKJ201205004.htm
    [12] Wang L, Zhong C X, Guo Y, Wu Y F, Guo J. Robust adaptive attitude control for flexible spacecraft in the presence of SGCMG friction nonlinearity. International Journal of Robust and Nonlinear Control, 2018, 28(9): 3324-3341 doi: 10.1002/rnc.4079
    [13] Cao X B, Wu B L. Indirect adaptive control for attitude tracking of spacecraft with unknown reaction wheel friction. Aerospace Science and Technology, 2015, 47: 493-500 doi: 10.1016/j.ast.2015.10.017
    [14] 李冬柏, 吴宝林, 张迎春. 考虑未知飞轮摩擦力矩的航天器姿态跟踪鲁棒自适应控制. 宇航学报, 2016, 37(2): 175-181 https://www.cnki.com.cn/Article/CJFDTOTAL-YHXB201602007.htm

    Li Dong-Bai, Wu Bao-Lin, Zhang Ying-Chun. Robust adaptive control for spacecraft attitude tracking with unknown friction torque of reaction wheel. Journal of Astronautics, 2016, 37(2): 175-181 https://www.cnki.com.cn/Article/CJFDTOTAL-YHXB201602007.htm
    [15] Niu G, Hu Q. Robust finite-time observer design for rigid spacecraft with reaction wheel friction. In: Proceedings of the 35th Chinese Control Conference. Chengdu, China: IEEE, 2016. 10679-10683
    [16] Heiberg C J, Bailey D, Wie B. Precision spacecraft pointing using single-gimbal control moment gyroscopes with disturbance. Journal of Guidance Control and Dynamics, 2000, 23(1): 77-85 doi: 10.2514/2.4489
    [17] MacKunis W, Leve F, Patre P M, Fitz-Coy N, Dixon W E. Adaptive neural network-based satellite attitude control in the presence of CMG uncertainty. Aerospace Science and Technology, 2016, 54: 218-228 doi: 10.1016/j.ast.2016.04.022
    [18] Hughes P C. Spacecraft Attitude Dynamics. New York: Courier Dover Publications, 2012. 93-129
    [19] Wie B. New singularity escape/avoidance steering logic for control moment gyro systems. Journal of Guidance, Control, and Dynamics, 2005, 28(5): 948-956 doi: 10.2514/1.10136
    [20] De Wit C C, Olsson H, Astrom K J, Lischinsky P. A new model for control of systems with friction. IEEE Transactions on Automatic Control, 1995, 40(3): 419-425 doi: 10.1109/9.376053
    [21] De Wit C C, Lischinsky P. Adaptive friction compensation with partially known dynamic friction model. International Journal of Adaptive Control and Signal Processing, 1997, 11(1): 65-80 doi: 10.1002/(SICI)1099-1115(199702)11:1%3C65::AID-ACS395%3E3.0.CO;2-3/abstract
    [22] Sidi M J. Spacecraft Dynamics and Control: A Practical Engineering Approach. Cambridge: Cambridge University Press, 1997.
    [23] Ma Y, Jiang B, Tao G, Cheng Y H. Uncertainty decomposition-based fault-tolerant adaptive control of flexible spacecraft. IEEE Transactions on Aerospace and Electronic Systems, 2015, 51(2): 1053-1068 http://smartsearch.nstl.gov.cn/paper_detail.html?id=711cece5fac30dddb4cb6cdbe14ba907
    [24] Hardy G H, Littlewood J E, Pólya G. Inequalities. Cambridge: Cambridge University Press, 1952.
    [25] Zhu Z, Xia Y Q, Fu M Y. Attitude stabilization of rigid spacecraft with finite-time convergence. International Journal of Robust and Nonlinear Control, 2011, 21(6): 686-702 doi: 10.1002/rnc.1624
    [26] Zou A M, Kumar K D, Ruiter A H J. Robust attitude tracking control of spacecraft under control input magnitude and rate saturations. International Journal of Robust and Nonlinear Control, 2015, 26(4): 799-815 doi: 10.1002/rnc.3338
    [27] Gennaro S D. Output stabilization of flexible spacecraft with active vibration suppression. IEEE Transactions on Aerospace and Electronic Systems, 2003, 39(3): 747-759 http://ieeexplore.ieee.org/document/1238733/references
    [28] Guo Y, Zhou C F, Yu Z, Chen Q W. Study on attitude control for move-to-rest manoeuvre of flexible spacecraft. International Journal of Modelling, Identification and Control, 2013, 19(1): 23-31 http://www.ingentaconnect.com/content/rsoc/17466172/2013/00000019/00000001/art00003
  • 期刊类型引用(7)

    1. 宋晓娟,范志文,吕书锋,岳宝增. 液体大幅晃动充液航天器分数阶滑模姿态机动控制. 宇航学报. 2024(07): 1146-1154 . 百度学术
    2. 袁利,姜甜甜. 航天器威胁规避智能自主控制技术研究综述. 自动化学报. 2023(02): 229-245 . 本站查看
    3. 管萍,黄明亮,戈新生,曹彧腾. 大挠性航天器的模糊模型预测控制. 航天控制. 2023(03): 43-49 . 百度学术
    4. 石永霞,胡庆雷,邵小东. 角速度受限下航天器姿态机动事件触发控制. 中国科学:信息科学. 2022(03): 506-520 . 百度学术
    5. 李新凯,张宏立,范文慧. 基于时变障碍李雅普诺夫函数的变体无人机有限时间控制. 自动化学报. 2022(08): 2062-2074 . 本站查看
    6. 管萍,吴希岩,戈新生. 基于Tube的挠性航天器模型预测姿态控制及主动振动控制. 振动与冲击. 2022(16): 261-270 . 百度学术
    7. 王龙龙,蓝鸿,周胥,陈度,何志祝,李臻. 拖拉机极限态侧翻回稳陀螺主动控制系统设计与试验. 农业机械学报. 2022(S2): 320-327 . 百度学术

    其他类型引用(6)

  • 加载中
  • 图(8) / 表(1)
    计量
    • 文章访问数:  840
    • HTML全文浏览量:  260
    • PDF下载量:  199
    • 被引次数: 13
    出版历程
    • 收稿日期:  2018-06-25
    • 录用日期:  2018-11-22
    • 刊出日期:  2021-04-02

    目录

    /

    返回文章
    返回