具有反馈信息的迭代学习控制律在Lebesgue-p范数意义下的收敛性

针对一类线性时不变系统, 提出了具有反馈信息的PD-型(Proportional-derivative-type)迭代学习控制律, 利用卷积的推广的Young不等式, 分析了控制律在Lebesgue-p范数意义下的单调收敛性. 分析表明, 收敛性不但决定于系统的输入输出矩阵和控制律的微分学习增益, 而且依赖于系统的状态矩阵和控制律的比例学习增益; 进一步, 当适当选取反馈增益时, 反馈信息可加快典型的PD-型迭代学习控制律的单调收敛性. 数值仿真验证了理论分析的正确性和控制律的有效性.     

Lebesgue-p范数意义下区间可调节的变增益加速迭代学习控制

为加快迭代学习控制律的收敛速度,针对线性时不变(LTI)系统,以PD-型学习律为例,提出一种区间可调节的具有指数加速的迭代学习控制算法.首先,根据每次学习效果确定下一次迭代需要修正的区间并在该区间内修正控制律增益;然后,在Lebesgue-p范数意义下分析所提出算法的收敛性并给出其收敛条件;最后,通过理论分析表明,收敛速度主要取决于被控对象、控制律增益、修正指数和学习区间的大小.在相同仿真条件下,与传统算法相比,所提出算法具有更快的收敛速度.     

正则系统在Lebesgue-p范数意义下的快速迭代学习控制

针对一类线性正则系统,传统迭代学习控制算法收敛速度较低的问题,设计了一种快速迭代学习控制算法。该算法在传统P型迭代学习控制算法基础上,增加了由相邻两次迭代时跟踪误差构成的上一次差分信号和当前差分信号,并在Lebesgue-p范数度量意义下,利用卷积推广的Young不等式严格证明了,当迭代次数趋于无穷大时,系统的跟踪误差收敛于零,并给出了算法的收敛条件。该算法与传统P型迭代学习控制算法相比,不仅提高了收敛速度,而且还避免了采用λ范数度量跟踪误差的缺陷,最后通过仿真结果进一步验证了所提算法的有效性。     

Lebesgue-?? 范数意义下对初态误差进行加速修正的迭代学习控制

针对一类多输入多输出线性时不变系统, 提出一种初态误差加速修正的PD-型迭代学习算法. 针对系统的任意初始状态, 在时间轴上设计一个随迭代次数增加而缩短的修正区间. 在该区间上, 控制算法对初始状态偏差进行修正; 修正区间外, 算法与无初始误差的学习律等同. 在Lebesgue-?? 范数度量跟踪误差意义下, 利用卷积的推 广Young 不等式证明了所提出学习控制律的收敛性. 数值仿真验证了该控制律的有效性.

     

反馈辅助PD型迭代学习控制:初值问题及修正策略

讨论迭代初态与期望初态存在固定偏移情形下 的迭代学习控制问题, 提出带有反馈辅助项的PD型迭代学习控制算法, 可实现系统输出对期望轨迹的渐近跟踪. 为了进一步实现输出轨迹在预定有限区间上对期望轨迹的完全跟踪, 提出分别带有初始修正作用和终态吸引的学习算法. 文中给出所提出的学习算法的极限轨迹, 并对学习算法进行收敛性分析, 推导出收敛性充分条件, 可用于学习增益的确定. 通过数值结果, 验证所提学习算法的有效性.     

线性时不变系统PID–型迭代学习控制律的单调收敛形态

传统的迭代学习控制机理中,积分补偿是典型的策略之一,但其跟踪效用并不明确.本文针对连续线性时不变系统,对传统的PD–型迭代学习控制律嵌入积分补偿,利用分部积分法和推广的卷积Young不等式,在Lebesguep范数意义下,理论分析一阶和二阶PID–型迭代学习控制律的收敛性态.结果表明,当比例、积分和导数学习增益满足适当条件时,一阶PID–型迭代学习控制律是单调收敛的,二阶PID–型迭代学习控制律是双迭代单调收敛的.数值仿真验证了积分补偿可有效地提高系统的跟踪性能.     

不确定离散线性系统的鲁棒单调反馈–前馈迭代学习控制

针对一类不确定离散线性系统,提出一种沿迭代方向鲁棒单调收敛和沿时间方向有界输入有界输出(bouned-input bounded-output,BIBO)稳定的反馈–前馈迭代学习控制策略.首先,将不确定反馈–前馈迭代学习系统表示为不确定二维Roesser模型系统;然后,把二维系统沿迭代方向的鲁棒单调收敛问题转化成一维系统的H∞干扰抑制控制问题,并给出系统的稳定性证明和用线性矩阵不等式(linear matrix inequality,LMI)表示的沿迭代方向鲁棒单调收敛的充分条件,该LMI充分条件不仅可以用于确定反馈–前馈控制器的增益矩阵,而且还可以保证系统沿时间轴方向是BIBO稳定的;最后,仿真结果证明了该反馈–前馈迭代学习控制策略的有效性.     

具有通信约束的反馈辅助PD型量化迭代学习控制

本文针对网络线性系统, 研究了具有通信约束的反馈辅助PD型迭代学习控制问题. 信号从远程设备传输到 迭代学习控制器过程中, 存在数据量化与数据包丢失的情况. 将数据包丢失模型描述为具有已知概率的伯努利二 进制序列, 采用扇形界方法处理数据量化误差, 提出了一种反馈辅助PD型迭代学习控制算法. 采用压缩映射法分析 证明了在存在数据量化和丢失的情况下, 所提控制算法依然可以保证跟踪误差渐近收敛到零. 并进一步对存在初 始状态偏移时所提算法的鲁棒性进行了讨论. 最后, 通过仿真示例, 对比验证了理论结果的有效性和优越性.     

具有初始状态不确定性的非线性系统脉冲补偿迭代学习控制

针对于具有初始状态不确定性的非线性时不变系统,采用矩形脉冲信号补偿传统的比例微分型一阶和二阶迭代学习控制律.在Lebesgue-p范数度量跟踪误差意义下,利用卷积的推广的Young不等式分析学习控制律的跟踪性能.分析表明,在适当选取比例学习增益,微分学习增益和非线性状态函数的Lipschitz常数以保证收敛因子小于1的前提下,渐近跟踪误差是由初始状态不确定性引起的,而且可通过调节补偿因子予以消减.数值仿真验证了补偿策略的有效性和理论分析的正确性.     

带有误差修正项的自适应学习控制

本文讨论了二阶系统在任意初态偏差下的自适应控制问题,借助学习控制及其初始修正的思想,提出了两种带有修正初态偏差功能的自适应控制策略:一阶吸引子控制器和零阶吸引子控制器.两种控制器都是将整个控制过程分成若干个等长时间的子过程,在每个子过程中控制算法都会进行误差校正和参数学习.其中,一阶吸引子控制器在每个子过程中同时修正所...     

Convergence properties concerning Lebesgue-p norm of iterative learning control for a class of fractional differential systems

This paper investigates the monotonic convergence and speed comparison of first- and second-order proportional-α-order-integral-derivative-type ( ${\mathit{PI}}^{\alpha }D-$ type) iterative learning control (ILC) schemes for a linear time-invariant (LTI) system, which is governed by the fractional differential equation with order $\alpha \in \left(1,2\right)$. By introducing the Lebesgue-p ( $L^P$) norm and utilizing the property of the Mittag-Leffler function and the boundedness feature of the fractional integration operator, the sufficient condition for the monotonic convergence of the first-order updating law is strictly analyzed. Therewith, the sufficient condition of the second-order learning law is established using the same means as the first one. The obtained results objectively reveal the impact of the inherent attributes of system dynamics and the constructive mode of the ILC rule on convergence. Based on the sufficient condition of first/second-order updating law, the convergence speed of first- and second-order schemes is determined quantitatively. The quantitative result demonstrates that the convergence speed of second-order law can be faster than the first one when the learning gains and weighting coefficients are properly selected. Finally, the effectiveness of the proposed methods is illustrated by the numerical simulations.     

Pulse compensation for PD-type iterative learning control against initial state shift

The rectangular pulse function is adopted to incorporate feed-forward compensation for various proportional-derivative-type iterative learning control updating laws applied to a class of linear time-invariant systems with initial state shift. The objective of pulse compensation is to suppress the tracking discrepancy incurred by initial state shift. By means of the generalised Young inequality of the convolution integral, the tracking performance of the pulse-based learning updating laws is analysed and the suppressive effect of the pulse compensation is evaluated by measuring the tracking error in the sense of Lebesgue-p norm. The derivation clarifies that the upper bound of the asymptotical tracking error can be improved by tuning the compensation gain properly though it is determined not only by the proportional and derivative learning gains but also by the system state, input and output matrices as well. Numerical simulations show that pulse compensation can effectively suppress the tracking error caused by initial state shift.     

Monotonic convergence of iterative learning control systems with variable pass length

A growing number of researchers consider iterative learning control (ILC) a promising tool for numerous control problems in biomedical application systems. We will briefly discuss why classical ILC theory is technically too restrictive for some of these applications. Subsequently, we will extend the classical ILC design in the lifted systems framework to the class of repetitive trajectory tracking tasks with variable pass length. We will analyse the closed-loop dynamics for two standard learning laws, and we will discuss in which sense the tracking error can be reduced by which controller design strategies. Necessary and sufficient conditions for monotonic convergence will be derived. We then summarise all results in a set of practical controller design guidelines. Finally, a simulation study is presented, which demonstrates the usefulness of these guidelines and illustrates the special dynamics that occur in variable pass length learning.     

A numerical method for determining monotonicity and convergence rate in iterative learning control

In iterative learning control (ILC), a lifted system representation is often used for design and analysis to determine the convergence rate of the learning algorithm. Computation of the convergence rate in the lifted setting requires construction of large N×N matrices, where N is the number of data points in an iteration. The convergence rate computation is O(N²) and is typically limited to short iteration lengths because of computational memory constraints. As an alternative approach, the implicitly restarted Arnoldi/Lanczos method (IRLM) can be used to calculate the ILC convergence rate with calculations of O(N). In this article, we show that the convergence rate calculation using IRLM can be performed using dynamic simulations rather than matrices, thereby eliminating the need for large matrix construction. In addition to faster computation, IRLM enables the calculation of the ILC convergence rate for long iteration lengths. To illustrate generality, this method is presented for multi-input multi-output, linear time-varying discrete-time systems.     

Monotonic convergence and robustness of higher‐order gain‐adaptive iterative learning control

For a class of repetitive linear discrete time‐invariant systems with higher relative degree, a higher‐order gain‐adaptive iterative learning control (HOGAILC) is developed while minimizing the energy increment of two adjacent tracking errors with the argument being the iteration‐time‐variable learning‐gain vector (ITVLGV). By taking advantage of rows/columns exchanging transformation of matrix, the ITVLGV is achieved in an explicit form which is dependent upon the system Markov parameters and adaptive to the iterationwise tracking‐error vector. Algebraic derivation demonstrates that the HOGAILC is strictly monotonously convergent. On the basis of the adaptive mode, a damping quasi‐HOGAILC strategy is exploited while the uncertainties of the system Markov parameters exist. Rigorous analysis delivers that the damping quasi‐scheme is strictly monotonically convergent and thus the HOGAILC mechanism is robust to a wider range of uncertainty of system parameters and the damping factor may relax the uncertainty range. Numerical simulations are made to illustrate the validity and the effectiveness.     

Multivariable norm optimal and parameter optimal iterative learning control: a unified formulation

This article investigates the two paradigms of norm optimal iterative learning control (NOILC) and parameter optimal iterative learning control (POILC) for multivariable (MIMO) ?-input, m-output linear discrete-time systems. The main result is a proof that, despite their algebraic and conceptual differences, they can be unified using linear quadratic multi-parameter optimisation techniques. In particular, whilst POILC has been naturally regarded as an approximation to NOILC, it is shown that the NOILC control law can be generated from a suitable choice of control law parameterisation and objective function in a multi-parameter MIMO POILC problem. The form of this equivalence is used to propose a new general approach to the construction of POILC problems for MIMO systems that approximates the solution of a given NOILC problem. An infinite number of such approximations exist. This great diversity is illustrated by the derivation of new convergent algorithms based on time interval and gradient partition that extend previously published work.