网络控制系统中迭代学习控制算法的鲁棒收敛性分析

Robust convergence of iterative learning control for networked control systems

针对网络控制系统中受输入扰动和初态干扰的非线性系统,提出了一种迭代学习控制算法,讨论了输出数据丢失情况下系统输出对期望轨迹的跟踪问题。采用一个概率已知的随机贝努利序列来描述网络控制系统中的数据丢失现象,即如果输出数据没有丢失,则利用跟踪误差来修正上一次迭代学习时的控制量,从而获得系统当前控制量;如果存在输出数据丢失现象,则系统当前控制量保持上一次迭代时的控制量不变。给出了保证算法收敛的充分条件。从理论上证明了系统满足给定的收敛条件时,在数据丢失网络环境下具有输入扰动、初态扰动的被控系统随迭代学习次数的增加,系统输出除初始点以外都能够收敛于期望轨迹。通过仿真算例进一步验证了所提方法的有效性。

This paper investigates an iterative learning control algorithm for nonlinear systems in the network model subject to input and initial state disturbances.In the case of output data loss,it deals with systems output tracking problems for the expected trajectory.Data loss phenomenon in networked control systems is described by using a random Bernoulli sequence with known probability.Based on available output data,tracking error is used to modify control amount of the last iteration to obtain the current control amount of the system.In case of output data loss,current control states are kept to be the same with the last iteration.The paper gives the sufficient condition of the convergence.For given convergence condition,with the number of iterative learning increasing,it is proved theoretically that system outputs other than initial point will converge to desired trajectory in the presence of network data loss and input and initial disturbance.Simulated results are provided to demonstrate the feasibility and effectiveness of the proposed scheme based on iterative learning control.