en
×

分享给微信好友或者朋友圈

使用微信“扫一扫”功能。
参考文献 1
许建新, 侯忠生. 学习控制的现状与展望[J]. 自动化学报, 2005, 31(6):943-952.
XUJ X, HOUZ S. On learning control: The state of the art and perspective[J]. Acta Automatica Sinica, 2005, 31(6):943-952.
参考文献 2
ARIMOTOS, KAWAMURAS, MIYAZAKIF. Bettering operation of robots by learning[J]. Journal of Intelligent & Robotic Systems, 1984, 1(2):123-140.doi:10.1002/rob.4620010203
参考文献 3
BRISTOWD, THARAYILM, ALLEYNEA G. A survey of iterative learning control[J]. IEEE Control Systems Magazine, 2006, 26(3):96-114.
参考文献 4
ISHIHARAT, ABE K, TAKEDAH. A discrete-time design of robust iterative learning controllers[J]. IEEE Transactions on Systems Man and Cybernetics Part B, 1992, 22(1):74-84.doi:10.1109/21.141312
参考文献 5
SAABS S, VOGTW G, MICKLEM H. Learning control algorithms for tracking ‘slowly’ varying trajectories[J]. IEEE Transactions on Systems Man and Cybernetics Part B, 1997, 27(4):657-670.doi:10.1109/3477.604109
参考文献 6
HEINZINGERG, FENWICKD, PADENB, et al. Stability of learning control with disturbances and uncertain initial conditions[J]. IEEE Transactions on Automatic Control, 1992, 37(1):110-114.doi:10.1109/9.109644
参考文献 7
WANGD W. Convergence and robustness of discrete time nonlinear systems with iterative learning control[J]. Automatica, 1998, 34(11):1445-1448.
参考文献 8
ARIMOTOS. Learning control theory for robotic motion[J]. International Journal of Adaptive Control and Signal Processing, 1990, 4(6): 543-564.doi:10.1002/acs.4480040610
参考文献 9
LEE H S, BIENZ. Study on robustness of iterative learning control with nonzero initial error[J]. International Journal of Control, 1996, 64(3):345-359.
参考文献 10
PARKP, BIENZ, HWANGD. A study on the robustness of a PID-type iterative learning controller against initial state error[J]. International Journal of Systems Science, 1999, 30(1):49-59.
参考文献 11
SUNM X, WANGD W. Initial condition issues on iterative learning control for nonlinear systems with time delay[J]. International Journal of Systems Science, 2001,32(11):1365-1375.
参考文献 12
李国军, 陈东杰, 韩一士. 带有初态误差的高阶多智能体系统一致性跟踪[J]. 应用数学学报, 2018, 41(2):156-171.
LIG J, CHEND J, HANY S. Consensus tracking of high-order multi-agent systems with initial state errors[J]. Acta Mathematicae Applocatae Sinica, 2018, 41(2): 156-171.
参考文献 13
LIX D, CHOWT W S, HO J K L, et al. Iterative learning control with initial rectifying action for nonlinear continuous systems[J]. IET Control Theory & Applications, 2009, 3(1):49-55.doi:10.1049/iet-cta:20070486
参考文献 14
LIX D, WEIY S. Iterative learning control for linear discrete-time systems with high relative degree under initial state vibration[J]. IET Control Theory & Applications, 2016, 10(10): 1115-1126.doi:10.1049/iet-cta.2015.0826
参考文献 15
MENGD Y, MOOREK L. Robust iterative learning control for nonrepetitive uncertain systems[J]. IEEE Transactions on Automatic Control, 2017, 62(2): 907-913.doi:10.1109/tac.2016.2560961
参考文献 16
XUJ X, TANY. A composite energy function-based learning control approach for nonlinear systems with time-varying parametric uncertainties[J]. IEEE Transactions on Automatic Control, 2002, 47(11): 1940-1945.doi:10.1109/tac.2002.804460
参考文献 17
TAYEBIA, CHIENC. A unified adaptive iterative learning control framework for uncertain nonlinear systems[J]. IEEE Transactions on Automatic Control, 2007, 52(10):1907-1913.
参考文献 18
CHIENC , HSU C, YAOC. Fuzzy system-based adaptive iterative learning control for nonlinear plants with initial state errors[J]. IEEE Transactions on Fuzzy Systems, 2004, 12(5): 724-732.doi:10.1109/tfuzz.2004.834806
参考文献 19
孙明轩. 有限时间迭代学习控制[J]. 系统科学与数学, 2010, 30(6): 733-741.
SUNM X. Finite-time iterative learning control[J]. Journal of Systems Science and Mathematical Sciences, 2010, 30(6): 733-741.
参考文献 20
谢华英, 孙明轩. 有限时间死区修正迭代学习控制器的设计[J]. 控制理论与应用, 2009, 26(11):1225-1231.
XIEH Y, SUNM X. Design of iterative learning controllers with finite-time dead-zone modification[J]. Control Theory & Applications, 2009, 26(11):1225-1231.
参考文献 21
齐丽强, 孙明轩,管海娃. 非参数不确定系统的有限时间迭代学习控制[J]. 自动化学报, 2014, 40(7): 1320-1327.
QIL Q, SUNM X, GUANH W. Finite-time iterative learning control for systems with nonparametric uncertainties[J]. Acta Automatica Sinica, 2014, 40(7):1320-1327.
参考文献 22
YINC K, XUJ X, HOUZ S. A high-order internal model based iterative learning control scheme for nonlinear systems with time-iteration-varying parameters[J]. IEEE Transactions on Automatic Control, 2010, 56(11): 2665-2670.
目录 contents

    摘要

    针对任意初态下带有扰动的线性定常系统, 提出了相应的控制算法。该算法将受控过程分成无穷个子过程, 系统利用上一子过程的输入输出信息来调节当前过程的输入,以期获得更好的控制效果。在控制过程中, 基于迭代学习控制思想, 借助初始修正手段, 可使系统的跟踪误差达到任意小, 且当系统无扰动时,在指定区间内可实现完全跟踪。最后, 通过仿真算例, 验证了算法的有效性。

    Abstract

    A control algorithm for the linear time-invariant system is presented in this paper. It divides the controlled process into infinite subprocesses. For each subprocess, the system uses the input and output information of the last subprocess to adjust the current input in order to achieve a better control effect. In the control process, based on the iterative learning control theory, this algorithm solves the control problem using the initial rectifying methods. It can make the tracking error of the system converge to an arbitrarily small range, and achieve complete tracking if there is no disturbance in the specified time period. Finally, the effectiveness of the stated algorithm is demonstrated by the simulation result.

    学习本是指人们通过阅读、听讲、思考、研究、实践等途径获得知识或技能的过程,目前已被广泛引申到其他领域。控制领域的学习大致可分为三类: 单一目标学习, 多目标学习以及量化的生物学习。经过研究者的不断努力, 学习控制理论体系已日益完善并成为控制方法论中的一个重要分支。学习控制主要包括迭代学习控制和重复学习控[1]。迭代学习控制(iterative learning control, 简称ILC)适用于在有限区间重复作业的场合,利用先前的控制输入和输出产生当前次的输入, 以便改进输出效果, 经过多次迭代以后, 系统能够实现完全无误差跟[2,3]。重复学习控制同样着眼于利用受控对象和目标轨迹的周期性或重复性, 利用上一周期的输入输出信息来改善当前的控制输入, 以期获得理想的输出效果。最大的区别在于迭代学习控制目标是在有限区间上达到完全跟踪,而重复学习控制目标是在无限区间上达到渐近跟踪。由于设计简单、 在线计算量小、控制效果好, 学习控制被应用于工业机器人控制、化工过程控制等场合。

    应用ILC方法时, 要求每次的初始定位必须处于某个理想位置, 但实际上无法做到。实际应用时只能在某个指定区间实现完全跟踪。在此情形下, 可放松对初始定位条件的限制, 即每次迭代时的初始状态可不相同。重复学习控制对初始定位操作不作要求,从而回避了迭代学习控制中的初始定位问题, 一定程度上推广了重复学习控制的应用范围。

    在有初态偏差的情形下, 为了实现完全跟踪, 研究者做了大量有益的工[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22]。目前的解决方法主要分为两类, 一类是压缩映射方法, 另一类是Lyapunov-like方法。LEE[9]和PARK[10]提出了一种基于固定初态误差的修正方法, 能在指定区间实现完全跟踪;针对线性连续多智能系统, 李国军[12]提出了一种用step-by-step控制器修正高阶多智能体的初态误差方法,但该方法只能修正固定初态误差, 无法修正任意初态误差;LI[13]提出了一种针对任意初态误差的修正方法, 但无法完全修正初始误差, 而且只适用于一阶系统。对于相对高阶离散系统, 在初态偏差满足某种特定条件下, LI[14]实现了精确跟踪; 对于迭代时变不确定系统, MEMG[15]实现了有界跟踪。对于任意初态误差的情形, 一般采用Lyapunov-like分析方[16,17,18,19,20,21,22];CHIEN[18]应用边界层方法达到了渐近跟踪;文献[19,20,21]利用吸引子达到了实际完全跟踪;YIN[22]提出了一种离散自适应ILC方法, 能确保系统沿着迭代轴渐近收敛、沿着时间轴在指定区间内逐点收敛。

    无论是迭代学习控制还是重复学习控制都有自己的优势。但一般的过程控制,既不满足迭代学习控制有限区间重复作业的要求,也不满足重复学习控制中目标轨迹具有周期性或重复性的特点。本文针对任意初态下带有扰动的线性定常系统, 利用迭代学习控制的思想, 提出了一种新的学习控制算法, 该算法将受控过程分成无穷个子过程,不仅可利用上一子过程的输入输出信息来调节当前的输入, 而且还可通过初始修正策略获得更快更好的控制效果。该算法系统稳定, 且当子过程长度趋向零时, 可实现任意小误差跟踪。

  • 1 问题阐述

    考虑线性定常系统

    x˙(t)=Ax(t)+Bu(t)+ω(t),y(t)=Cx(t),
    (1)

    其中,t[0,),x(t)Rn,y(t)Rr,u(t)Rm分别表示系统状态、控制输入和输出向量,ω(t)Rn为系统扰动。A,B,C是合适维数的系统参数矩阵,并且B,C可逆。

    yd(t)是给定的期望轨迹,xd(t)是给定的期望状态,e(t)=yd(t)-y(t)表示输出误差。

    假设 初始状态x(0)随机,并且状态x(t)可测。

    为便于收敛性分析, 采用以下形式来描述系统(1):

    x˙k(t)=Axk(t)+Buk(t)+ωk(t),yk(t)=Cxk(t),
    (2)

    其中,t[0,h]指有限时间, h为预设的时间常量;k=0,1,2,,可以理解为迭代次数;xk(t)Rn,yk(t)Rr,uk(t)Rm,ωk(t)Rn(相应可以简写为xk,yk,uk,ωk)分别表示系统状态、控制输入、 输出和扰动向量。

    yk,d(t)xk,d(t)(可简写为yk,dxk,d)分别表示kh+t时刻的期望轨迹和期望状态,ek(t)=yk,d(t)-yk(t)表示kh+t时刻的输出误差。

    由系统(2)和系统(1),有

    xk(t)=x(kh+t),xk(h)=xk+1(0)
    (3)

    同理,yk(t),yk,d(t),xk,d(t),ωk(t)ek(t)也具有类似于式(3)的性质。

  • 2 控制器设计

    本节的任务是设计一套控制算法, 使得系统(2)能跟踪期望轨迹。为此, 提出下面的控制律:

    uk+1(t)=uk(t)+Γ1e˙k+1(t)+Γ0e˙k(t)-Θ(t)Ξk(0)+Θ(t)Ξk+1(0)-B-1A[xk+1(t)-xk(t)]+B-1[x˙k+1,d(t)-x˙k,d(t)]
    (4)

    其中,

    Θ(t)=π2hsinπ(t-kh)h,t[kh,(k+1)h),Ξk(0)=(CB)-1(I-CBΓ0)ek(0),Ξk+1(0)=(CB)-1(I+CBΓ1)ek+1(0)

    事实上, 控制律中的函数Θ(t)并不唯一, 但须满足以下性质:

    性质 对函数Θ(t), 有

    I=kh(k+1)hΘ(τ)dτ=1

    注1此性质表明,Θ(t)函数经过1次积分后具有与脉冲函数类似的作用。而实际中, 脉冲函数是不存在的, 因此可用此函数代替脉冲函数。

  • 3 收敛性分析

    这一节着重分析系统(2)在应用控制律(4)以后的收敛性。上节提出的控制律(4)中, 引入了初始修正函数Θ(t), 关于初始修正函数有以下定理:

    定理 如果初始状态是任意有限值,C,B,I+CBΓ1都是可逆矩阵, 并且满足

    (I+CBΓ1)-1(I-CBΓ0)1
    (5)

    那么,修正控制律(4)能使系统(2)跟踪目标轨迹, 并且当k时, 跟踪误差可达任意小。

    证明 为书写方便, 记

    Δuk(t)=uk+1(t)-uk(t),Δxk(t)=xk+1(t)-xk(t),Δωk(t)=ωk+1(t)-ωk(t),

    t[kh,(k+1)h)时, 可得

    xk+1(t)-xk(t)=[xk+1(0)-xk(0)]+0tΔx˙k(τ)dτ=[xk+1(0)-xk(0)]+BΓ0[ek(t)-ek(0)]+BΓ1[ek+1(t)-ek+1(0)]+0tΔωk(τ)dτ+0tBΘ(τ)[Ξk+1(0)-Ξk(0)]dτ+0t[x˙k+1,d(τ)-x˙k,d(τ)]dτ=[xk+1(0)-xk(0)]+BΓ0[ek(t)-ek(0)]+BΓ1[ek+1(t)-ek+1(0)]+0tΔωk(τ)dτ+0tBΘ(τ)[Ξk+1(0)-Ξk(0)]dτ+xk+1,d(t)-xk,d(t)-[xk+1,d(0)-xk,d(0)],

    上式两端同乘以矩阵C,整理后可得

    (I-CBΓ0)ek(t)-(I-CBΓ0)ek(0)+0tCBΘ(τ)Ξk(0)dτ=(I+CBΓ1)ek+1(t)-(I+CBΓ1)ek+1(0)+0tCBΘ(τ)Ξk+1(0)dτ+0tCΔωk(τ)dτ,

    代入Ξk(0)Ξk+1(0)的表达式, 进一步化简上式,可得

    (I-CBΓ0)ek(t)-(I-CBΓ0)ek(0)+0tΘ(τ)(I-CBΓ0)ek(0)dτ=(I+CBΓ1)[ek+1(t)-ek+1(0)+0tΘ(τ)ek+1(0)dτ]+0tCΔωk(τ)dτ,

    上式两端同时左乘(I+CBΓ1)-1,可得

    (I+CBΓ1)-1(I-CBΓ0)[ek(t)-ek(0)+0tΘ(τ)ek(0)dτ]=ek+1(t)-ek+1(0)+0tΘ(τ)ek+1(0)dτ+(I+CBΓ1)-10tCΔωk(τ)dτ
    (6)

    假设轨迹

    yk,d*(t)=yk,d(t)+ek(0)-0tΘ(τ)ek(0)dτ

    可实现, 相应的误差记为ek*(t)=yk,d*(t)-yk(t), 而且

    ek*(t)=ek(t)+yk,d*(t)-yk,d(t),
    (7)

    由式(6)和(7)得

    ek+1*(t)=(I+CBΓ1)-1(I-CBΓ0)ek*(t)-(I+CBΓ1)-10tCΔωk(τ)dτ

    若记β=max{||(I+CBΓ1)-10tCΔωk(τ)dτ||,k=1,2,},则当满足条件

    (I+CBΓ1)-1(I-CBΓ0)1

    时, 有

    limkek+1*(t)β(I+CBΓ1)-1(I-CBΓ0),

    且当t=h时, 根据Θ(t)函数的性质有

    limkek+1(t)=limkek+1*(t)β(I+CBΓ1)-1(I-CBΓ0)
    (8)

    由式(7)有

    ek+1(t)=ek+1*(t)-ek+1(0)+0tΘ(τ)ek+1(0)d(τ)=ek+1*(t)-ek*(h)+0tΘ(τ)ek*(h)d(τ)

    由上面的推导结果可知,ek+1*(t)有界, 所以ek+1(t)有界, 因此在控制律(4)的作用下, 系统(2)稳定。

    注2 由β的定义知,limh0β=0

    事实上, 如果系统(2)无扰动, 只需采用以下控制律即可:

    u(t)=Θ(t)Ξ(0)-B-1A[x(t)-x˙d(t)],

    其中,

    Θ(t)=π2hsin(πth),t[0,h],0,t(h,),
    Ξ(0)=(CB)-1e(0)

    在此情形下,

    x(t)=0tx˙(τ)dτ+x(0)=Ax(τ)+BΘ(τ)[Ξ0-[Ax(τ)-x˙dτ]dτ+x(0)=0tBΘ(τ)Ξ(0)dτ+xd(t)-xd(0)+x(0)

    上式两端同乘以C并代入Ξ(0), 化简后可得

    e(t)=e(0)-0tΘ(τ)e(0)dτ

    因此,当th时, 误差e(t)0

  • 4 数值仿真

    考虑连续系统

    x(t)=10.501x(t)+0.20.50.10.5u(t)+rand,y(t)=10.30.21x(t),

    显然, 该系统不稳定。其作业区间为[0,20], 系统的跟踪轨迹为yd,1(t)=cos(0.1πt),yd,2(t)=sin(0.1πt)。系统扰动为rand(rand表示0~1之间的随机数),系统运行之前的初始状态为x1(0)=0,x2(0)=0。 控制增益Γ0=0.8,Γ1=0.8, 修正时间h=0.2。仿真结果如图1,2,3所示。

    图1
                            期望轨迹yd(t)和实际轨迹y(t)

    图1 期望轨迹yd(t)和实际轨迹y(t)

    Fig.1 Desired trajectory yd(t)and actual trajectory y(t)

    图 2
                            实际跟踪误差e(t)

    图 2 实际跟踪误差e(t)

    Fig.2 Actual tracking error e(t)

    图 3
                            控制量u(t)

    图 3 控制量u(t)

    Fig.3 Actual tracking error u(t)

    从图1和图2中可以看出,系统在初始阶段并未跟踪上目标轨迹,而是经过若干个周期的修正后才实现跟踪,但仍无法达到完全跟踪。

    从图3中可以看出, 系统以0.2为周期不断进行修正, 直至误差为零。由于跟踪过程中存在随机扰动, 控制器存在轻微颤振。

    若系统无扰动, 则只需经过1次修正即可达到完全跟踪。

  • 5 结论

    讨论了带有扰动的线性定常系统的学习控制问题, 基于迭代学习控制思想, 借助压缩映射手段, 提出了一种带修正因子的控制策略。 对于连续系统, 通过不断修正跟踪误差,将误差控制在一定范围, 当修正区间长度趋于零时, 可实现任意小误差跟踪。特别地, 当系统无扰动时, 可在指定区间实现完全跟踪。 本文从理论和实践两方面验证了算法的有效性。

  • 参考文献(References)

    • 1

      许建新, 侯忠生. 学习控制的现状与展望[J]. 自动化学报, 2005, 31(6):943-952.

      XU J X, HOU Z S. On learning control: The state of the art and perspective[J]. Acta Automatica Sinica, 2005, 31(6):943-952.

    • 2

      ARIMOTO S, KAWAMURA S, MIYAZAKI F. Bettering operation of robots by learning[J]. Journal of Intelligent & Robotic Systems, 1984, 1(2):123-140.doi:10.1002/rob.4620010203

    • 3

      BRISTOW D, THARAYIL M, ALLEYNE A G. A survey of iterative learning control[J]. IEEE Control Systems Magazine, 2006, 26(3):96-114.

    • 4

      ISHIHARA T, ABE K, TAKEDA H. A discrete-time design of robust iterative learning controllers[J]. IEEE Transactions on Systems Man and Cybernetics Part B, 1992, 22(1):74-84.doi:10.1109/21.141312

    • 5

      SAAB S S, VOGT W G, MICKLE M H. Learning control algorithms for tracking ‘slowly’ varying trajectories[J]. IEEE Transactions on Systems Man and Cybernetics Part B, 1997, 27(4):657-670.doi:10.1109/3477.604109

    • 6

      HEINZINGER G, FENWICK D, PADEN B, et al. Stability of learning control with disturbances and uncertain initial conditions[J]. IEEE Transactions on Automatic Control, 1992, 37(1):110-114.doi:10.1109/9.109644

    • 7

      WANG D W. Convergence and robustness of discrete time nonlinear systems with iterative learning control[J]. Automatica, 1998, 34(11):1445-1448.

    • 8

      ARIMOTO S. Learning control theory for robotic motion[J]. International Journal of Adaptive Control and Signal Processing, 1990, 4(6): 543-564.doi:10.1002/acs.4480040610

    • 9

      LEE H S, BIEN Z. Study on robustness of iterative learning control with nonzero initial error[J]. International Journal of Control, 1996, 64(3):345-359.

    • 10

      PARK P, BIEN Z, HWANG D. A study on the robustness of a PID-type iterative learning controller against initial state error[J]. International Journal of Systems Science, 1999, 30(1):49-59.

    • 11

      SUN M X, WANG D W. Initial condition issues on iterative learning control for nonlinear systems with time delay[J]. International Journal of Systems Science, 2001,32(11):1365-1375.

    • 12

      李国军, 陈东杰, 韩一士. 带有初态误差的高阶多智能体系统一致性跟踪[J]. 应用数学学报, 2018, 41(2):156-171.

      LI G J, CHEN D J, HAN Y S. Consensus tracking of high-order multi-agent systems with initial state errors[J]. Acta Mathematicae Applocatae Sinica, 2018, 41(2): 156-171.

    • 13

      LI X D, CHOW T W S, HO J K L, et al. Iterative learning control with initial rectifying action for nonlinear continuous systems[J]. IET Control Theory & Applications, 2009, 3(1):49-55.doi:10.1049/iet-cta:20070486

    • 14

      LI X D, WEI Y S. Iterative learning control for linear discrete-time systems with high relative degree under initial state vibration[J]. IET Control Theory & Applications, 2016, 10(10): 1115-1126.doi:10.1049/iet-cta.2015.0826

    • 15

      MENG D Y, MOORE K L. Robust iterative learning control for nonrepetitive uncertain systems[J]. IEEE Transactions on Automatic Control, 2017, 62(2): 907-913.doi:10.1109/tac.2016.2560961

    • 16

      XU J X, TAN Y. A composite energy function-based learning control approach for nonlinear systems with time-varying parametric uncertainties[J]. IEEE Transactions on Automatic Control, 2002, 47(11): 1940-1945.doi:10.1109/tac.2002.804460

    • 17

      TAYEBI A, CHIEN C. A unified adaptive iterative learning control framework for uncertain nonlinear systems[J]. IEEE Transactions on Automatic Control, 2007, 52(10):1907-1913.

    • 18

      CHIEN C , HSU C, YAO C. Fuzzy system-based adaptive iterative learning control for nonlinear plants with initial state errors[J]. IEEE Transactions on Fuzzy Systems, 2004, 12(5): 724-732.doi:10.1109/tfuzz.2004.834806

    • 19

      孙明轩. 有限时间迭代学习控制[J]. 系统科学与数学, 2010, 30(6): 733-741.

      SUN M X. Finite-time iterative learning control[J]. Journal of Systems Science and Mathematical Sciences, 2010, 30(6): 733-741.

    • 20

      谢华英, 孙明轩. 有限时间死区修正迭代学习控制器的设计[J]. 控制理论与应用, 2009, 26(11):1225-1231.

      XIE H Y, SUN M X. Design of iterative learning controllers with finite-time dead-zone modification[J]. Control Theory & Applications, 2009, 26(11):1225-1231.

    • 21

      齐丽强, 孙明轩,管海娃. 非参数不确定系统的有限时间迭代学习控制[J]. 自动化学报, 2014, 40(7): 1320-1327.

      QI L Q, SUN M X, GUAN H W. Finite-time iterative learning control for systems with nonparametric uncertainties[J]. Acta Automatica Sinica, 2014, 40(7):1320-1327.

    • 22

      YIN C K, XU J X, HOU Z S. A high-order internal model based iterative learning control scheme for nonlinear systems with time-iteration-varying parameters[J]. IEEE Transactions on Automatic Control, 2010, 56(11): 2665-2670.

李国军

机 构:浙江警察学院 公共基础部,浙江 杭州 310053

Affiliation:Basic Courses Department, Zhejiang Police College, Hangzhou 310053, China

邮 箱:liguojun@zjjcxy.cn.

作者简介:李国军(1979—), ORCID:https://orcid.org/0000-0002-1338-6331,男, 博士研究生, 讲师, 主要从事学习控制研究,E-mail:liguojun@zjjcxy.cn.

周国民

机 构:浙江警察学院 计算机与信息技术系,浙江 杭州 310053

Affiliation:Department of Computer and Information Technology, Zhejiang Police College, Hangzhou 310053, China

陈东杰

机 构:浙江警察学院 公共基础部,浙江 杭州 310053

Affiliation:Basic Courses Department, Zhejiang Police College, Hangzhou 310053, China

许中石

机 构:浙江警察学院 公共基础部,浙江 杭州 310053

Affiliation:Basic Courses Department, Zhejiang Police College, Hangzhou 310053, China

1008-9497-2019-46-3-328/alternativeImage/142b0790-5b54-4a48-b408-0be7ea2f97a3-F001.jpg
1008-9497-2019-46-3-328/alternativeImage/142b0790-5b54-4a48-b408-0be7ea2f97a3-F002.jpg
1008-9497-2019-46-3-328/alternativeImage/142b0790-5b54-4a48-b408-0be7ea2f97a3-F003.jpg

图1 期望轨迹yd(t)和实际轨迹y(t)

Fig.1 Desired trajectory yd(t)and actual trajectory y(t)

图 2 实际跟踪误差e(t)

Fig.2 Actual tracking error e(t)

图 3 控制量u(t)

Fig.3 Actual tracking error u(t)

image /

无注解

无注解

无注解

  • 参考文献(References)

    • 1

      许建新, 侯忠生. 学习控制的现状与展望[J]. 自动化学报, 2005, 31(6):943-952.

      XU J X, HOU Z S. On learning control: The state of the art and perspective[J]. Acta Automatica Sinica, 2005, 31(6):943-952.

    • 2

      ARIMOTO S, KAWAMURA S, MIYAZAKI F. Bettering operation of robots by learning[J]. Journal of Intelligent & Robotic Systems, 1984, 1(2):123-140.doi:10.1002/rob.4620010203

    • 3

      BRISTOW D, THARAYIL M, ALLEYNE A G. A survey of iterative learning control[J]. IEEE Control Systems Magazine, 2006, 26(3):96-114.

    • 4

      ISHIHARA T, ABE K, TAKEDA H. A discrete-time design of robust iterative learning controllers[J]. IEEE Transactions on Systems Man and Cybernetics Part B, 1992, 22(1):74-84.doi:10.1109/21.141312

    • 5

      SAAB S S, VOGT W G, MICKLE M H. Learning control algorithms for tracking ‘slowly’ varying trajectories[J]. IEEE Transactions on Systems Man and Cybernetics Part B, 1997, 27(4):657-670.doi:10.1109/3477.604109

    • 6

      HEINZINGER G, FENWICK D, PADEN B, et al. Stability of learning control with disturbances and uncertain initial conditions[J]. IEEE Transactions on Automatic Control, 1992, 37(1):110-114.doi:10.1109/9.109644

    • 7

      WANG D W. Convergence and robustness of discrete time nonlinear systems with iterative learning control[J]. Automatica, 1998, 34(11):1445-1448.

    • 8

      ARIMOTO S. Learning control theory for robotic motion[J]. International Journal of Adaptive Control and Signal Processing, 1990, 4(6): 543-564.doi:10.1002/acs.4480040610

    • 9

      LEE H S, BIEN Z. Study on robustness of iterative learning control with nonzero initial error[J]. International Journal of Control, 1996, 64(3):345-359.

    • 10

      PARK P, BIEN Z, HWANG D. A study on the robustness of a PID-type iterative learning controller against initial state error[J]. International Journal of Systems Science, 1999, 30(1):49-59.

    • 11

      SUN M X, WANG D W. Initial condition issues on iterative learning control for nonlinear systems with time delay[J]. International Journal of Systems Science, 2001,32(11):1365-1375.

    • 12

      李国军, 陈东杰, 韩一士. 带有初态误差的高阶多智能体系统一致性跟踪[J]. 应用数学学报, 2018, 41(2):156-171.

      LI G J, CHEN D J, HAN Y S. Consensus tracking of high-order multi-agent systems with initial state errors[J]. Acta Mathematicae Applocatae Sinica, 2018, 41(2): 156-171.

    • 13

      LI X D, CHOW T W S, HO J K L, et al. Iterative learning control with initial rectifying action for nonlinear continuous systems[J]. IET Control Theory & Applications, 2009, 3(1):49-55.doi:10.1049/iet-cta:20070486

    • 14

      LI X D, WEI Y S. Iterative learning control for linear discrete-time systems with high relative degree under initial state vibration[J]. IET Control Theory & Applications, 2016, 10(10): 1115-1126.doi:10.1049/iet-cta.2015.0826

    • 15

      MENG D Y, MOORE K L. Robust iterative learning control for nonrepetitive uncertain systems[J]. IEEE Transactions on Automatic Control, 2017, 62(2): 907-913.doi:10.1109/tac.2016.2560961

    • 16

      XU J X, TAN Y. A composite energy function-based learning control approach for nonlinear systems with time-varying parametric uncertainties[J]. IEEE Transactions on Automatic Control, 2002, 47(11): 1940-1945.doi:10.1109/tac.2002.804460

    • 17

      TAYEBI A, CHIEN C. A unified adaptive iterative learning control framework for uncertain nonlinear systems[J]. IEEE Transactions on Automatic Control, 2007, 52(10):1907-1913.

    • 18

      CHIEN C , HSU C, YAO C. Fuzzy system-based adaptive iterative learning control for nonlinear plants with initial state errors[J]. IEEE Transactions on Fuzzy Systems, 2004, 12(5): 724-732.doi:10.1109/tfuzz.2004.834806

    • 19

      孙明轩. 有限时间迭代学习控制[J]. 系统科学与数学, 2010, 30(6): 733-741.

      SUN M X. Finite-time iterative learning control[J]. Journal of Systems Science and Mathematical Sciences, 2010, 30(6): 733-741.

    • 20

      谢华英, 孙明轩. 有限时间死区修正迭代学习控制器的设计[J]. 控制理论与应用, 2009, 26(11):1225-1231.

      XIE H Y, SUN M X. Design of iterative learning controllers with finite-time dead-zone modification[J]. Control Theory & Applications, 2009, 26(11):1225-1231.

    • 21

      齐丽强, 孙明轩,管海娃. 非参数不确定系统的有限时间迭代学习控制[J]. 自动化学报, 2014, 40(7): 1320-1327.

      QI L Q, SUN M X, GUAN H W. Finite-time iterative learning control for systems with nonparametric uncertainties[J]. Acta Automatica Sinica, 2014, 40(7):1320-1327.

    • 22

      YIN C K, XU J X, HOU Z S. A high-order internal model based iterative learning control scheme for nonlinear systems with time-iteration-varying parameters[J]. IEEE Transactions on Automatic Control, 2010, 56(11): 2665-2670.