Adaptive Learning FCS-MPC for Power Converters

View at publisher Summarise with VIRAT
Authors: Xing Liu, Lin Qiu, Dengyin Jiang, Youtong Fang, Kui Wang, Yongdong Li, Jose Rodriguez
Cited by: 0
References: 45

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Toward bridging classical finite control-set model predictive control (FCS-MPC) and artificial intelligence technique, online learning predictive control has recently been proposed as a control design solution. Maximizing the performance of a power converter system is a challenging problem for learning-based FCS-MPC tasks, especially when low switching frequency requirements are mandatory. Motivated by this need, this article concentrates on a novel adaptive online learning robust predictive control architecture that aims to enhance the system performance by combining a gradient descent-barzilai borwein-based neural network control technique with a sequential FCS-MPC approach in the presence of model parametric uncertainties under low switching frequency scenario. The key characteristics of this established control architecture are that, first, the issue of robustness inherent in the classical FCS-MPC method is explicitly dealt with, second, the high switching frequency is alleviated, third, the parameters and learning rates of neural network are updated online, and, fourth, the need for weighting factors is eliminated. Finally, we demonstrate the presented methodology in contrast to existing FCS-MPC schemes via multiple benchmark cases from different literature, and the results are promising and inspire further investigation in this field.

Fields of Study & Subjects

Multilevel Inverters and Converters Advanced Control Systems Optimization Advanced DC-DC Converters

Publication details

Journal / source
IEEE Journal of Emerging and Selected Topics in Power Electronics
Volume
14
Issue
4
Pages
5289–5302
Publication date
25 March 2026
Language
EN