.png){kind=icon}
.png){kind=icon}
Return
Data-driven control of LTV systems with application to ship autopilot
.png){kind=icon}
PRE
AI
J
T
J
DOI：10.1016/j.conengprac.2026.107126.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
This paper addresses the data-driven control (DDC) of discrete-time linear time-varying (LTV) systems, where system matrices are unknown and disturbances are bounded. Using pre-collected input-state-output data, we design a moving-horizon controller that, at each time step, synthesizes a state-feedback law via a data-dependent optimization. The formulation minimizes a computable upper bound on the worst-case performance while ensuring internal stability and a specified H∞ performance level. We prove recursive feasibility, guaranteeing that the controller can be continuously implemented online. The proposed approach is validated on a ship autopilot system through both simulations and field experiments, demonstrating its robustness and practical feasibility.
Journal
IF:
4.6
Papers:
5.6K
Citations:
1.1W
Organization
Cited Papers