.png){kind=icon}
.png){kind=icon}
Return
Low-Complexity Double-Layered Iterative Learning Control for Nonlinear MIMO System Under Cyberattacks
.png){kind=icon}
PRE
AI
DOI：10.1109/TCYB.2025.3634747.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In this article, the double-layered iterative learning control (DLILC) approach is adopted to investigate the tracking control problem of repetitive nonlinear multiple-input–multiple-output (MIMO) systems under false data injection (FDI) attacks. Based on historical data, two control loops in the scheme are devised to improve tracking accuracy. More specifically, an outer loop adaptive set-point tuning mechanism is developed, which is independent of the inner-loop controller. Such a mechanism dynamically optimizes learning gains by leveraging historical data and significantly reduces reliance on preset system parameters. In the inner loop, a proportional-derivative controller is employed to form the feedback circuit. Furthermore, the double dynamic linearization technique is adopted to transform complex nonlinearities, coupling effects, and unknown uncertainties into a set of linearly estimable parameters. To address FDI attacks, an output observer-based real-time compensator is constructed, which is capable of promptly mitigating the impact of such attacks on system outputs. Simulation results demonstrate that the proposed scheme ensures high-precision tracking, substantially reduces computational burden, and exhibits superior resilience against attacks. The approach thus provides a new pathway toward secure and efficient iterative learning control of nonlinear systems.
Keywords:
Cyberattacks
data-driven control
double-layered learning control mechanism
iterative learning control
multiple-input–multiple-output (MIMO) systems
Journal
IF:
10.5
Papers:
1.1W
Citations:
5.0W
Organization
View more organizations