一种迭代未知输入观测器设计  

An Iterative Unknown Input Observer Design

作　　者：朱芳来[1] 张永军[1] 

机构地区：[1]同济大学电子与信息工程学院,上海201804

出　　处：《同济大学学报（自然科学版）》2014年第8期1251-1255,1266,共6页Journal of Tongji University:Natural Science

基　　金：国家自然科学基金(61074009);高等学校博士学科点专项科研基金(20110072110015);广西制造系统与先进制造技术重点实验室基金(PF110289);上海重点学科项目(B004)

摘　　要：针对具有未知输入的线性系统讨论了基于迭代控制思想的未知输入观测器设计方法.针对每一步迭代,由于当前未知输入已知,因而可以通过设计常规的Luenberger观测器得到当前步的状态估计和输出估计.在此基础上采用D-型迭代学习控制的思想,通过当前输出步误差和未知输入估计值提出下一步的未知输入的迭代估计方法,并利用此次未知输入估计值再次设计Luenberger观测器估计下一步的状态.之后,对迭代算法的收敛性进行了分析,并给出了迭代算法收敛的充分条件.最后,对一个实际模型进行仿真,验证了算法的有效性.Based on the thought of iterative control,this paper presents a method to design an unknown input observer for linear continuous-time systems with unknown input.In each step of iteration,the current input of iteration is known,so the traditional Luenberger observer is used to estimate the current system state and corresponding output.Based on D-type iterative learning control algorithm,the iterative method to estimate the unknown input of next step is proposed by using the current output error and estimated input,and the Luenberger observer is redesigned by using the estimated value to estimate the next step system state.The convergence of the iterative algorithm is analyzed,and the sufficient condition for its convergence is given.Finally,a numerical simulation of a practical model is given to illustrate the effectiveness of the proposed methods.

关 键 词：未知输入观测器 未知输入重构 D-型迭代学习控制 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]