Open AccessData‐driven tuning of feedforward controller structured with infinite impulse response filter via iterative learning control
Author(s) -
Zhang Xinxin,
Li Min,
Ding Huafeng,
Yao Xiangyu
Publication year - 2019
Publication title -
iet control theory and applications
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.059
H-Index - 108
eISSN - 1751-8652
pISSN - 1751-8644
DOI - 10.1049/iet-cta.2018.5616
Subject(s) - feed forward , control theory (sociology) , iterative learning control , infinite impulse response , computer science , tracking error , finite impulse response , controller (irrigation) , filter (signal processing) , control engineering , algorithm , engineering , digital filter , artificial intelligence , control (management) , agronomy , computer vision , biology
Iterative learning control (ILC) is an effective approach for tracking control system that performs repeating tasks. However, the performance of ILC is significantly deteriorated when the reference is changed. To obtain high tracking performance for both repeating and varying tasks, a novel data‐driven tuning method of feedforward controller structured with infinite impulse response (IIR) filter via ILC is developed in this study. Global optimal parameters of the feedforward controller are obtained by linear least‐squares method based on the optimal feedforward control force obtained by ILC, while model information is not required. Additionally, to deal with the possible instability problem of the feedforward controller structured with IIR filter, a stable approximation approach on the basis of zero‐phase‐error tracking algorithm is presented. The stable approximation approach can convert the approximation problem to a convex optimisation problem. Finally, the proposed approach is compared with the standard ILC and a data‐driven feedforward control structured with finite impulse response filter by two simulation studies. Simulation results demonstrate that the proposed data‐driven feedforward tuning method can achieve high tracking performance and is insensitive to reference variations.