.png){kind=icon}
.png){kind=icon}
Return
Linear-Parameterization-Based Model Free Adaptive Predictive Control
.png){kind=icon}
PRE
AI
DOI：10.1109/TASE.2025.3613545.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
A novel data-driven predictive control scheme is proposed for unknown nonlinear discrete-time systems. The control increment vector is linearly parameterized through the dynamic linearization (DL) on the ideal controllers along the prediction horizon, leading to a predictive control law. The control gain matrix is adaptively tuned through the recursive least square method. The distinctive predictive control scheme is purely data-driven, because the predictive control law does not rely on underlying system dynamics and the control gain matrix is tuned by using only the input/output data of the controlled system. The asymptotic convergence of the designed scheme is rigorously guaranteed under reasonable conditions. One numerical example and one welding example are simulated to verify the validity and applicability of the designed scheme.Note to Practitioners—This paper was motivated by the industrial problem of sending what control input to the machine to follow the given signals N steps in advance. This problem is applied in various fields such as autonomous driving vehicles and welding process. Existing methods require the system dynamics to be known, which is challenging. This paper suggests a new control method by using only the data generated from the industrial process without knowing the system dynamics. In this paper, we mathematically characterize the mechanics of the new method and its reliability. Preliminary mathematical simulation and welding process simulation suggest that this method is feasible but this method has not been incorporated with disturbances and constraints. In future research, we will consider the disturbances and constraints in the discussed method.
Keywords:
Dynamic linearization technique
nonlinear system
data-driven control
predictive control
adaptive control
Journal
IF:
6.4
Papers:
4.9K
Citations:
1.6W
View more organizations