A novel iterative learning control scheme based on Broyden‐class optimization method

Jiaxi Qian,Dong Shen
DOI: https://doi.org/10.1002/rnc.6972
IF: 3.8973
2023-09-02
International Journal of Robust and Nonlinear Control
Abstract:Summary This study proposes a novel iterative learning control scheme for discrete‐time linear systems based on the Broyden‐class optimization method. To overcome the difficulty of lacking system information, a cost function is introduced for the performance index by constructing a positive‐definite matrix with little system information. An optimization‐based learning control algorithm is proposed using a Hessian matrix approximation and the generated input sequence is demonstrated to exhibit a superlinear convergence rate. The proposed scheme is extended to address the point‐to‐point tracking problem. Numerical simulations are provided to verify the effectiveness of the proposed approach.
automation & control systems,engineering, electrical & electronic,mathematics, applied
What problem does this paper attempt to address?