Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Physics-Infused Neural MPC of a DC-DC Boost Converter with Adaptive Transient Recovery and Enhanced Dynamic Stability

Tahmin Mahmud

Abstract

DC-DC boost converters require advanced control to ensure efficiency and stability under varying loads. Traditional model predictive control (MPC) and data-driven neural network methods face challenges such as high complexity and limited physical constraint enforcement. This paper proposes a hybrid physics-informed neural network (PINN) combined with finite control set MPC (FCS-MPC) for boost converters. The PINN embeds physical laws into neural training, providing accurate state predictions, while FCS-MPC ensures constraint satisfaction and multi-objective optimization. The method features adaptive transient recovery, explicit duty-ratio control, and enhanced dynamic stability. Experimental results on a commercial boost module demonstrate improved transient response, reduced voltage ripple, and robust operation across conduction modes. The proposed framework offers a computationally efficient, physically consistent solution for real-time control in power electronics.

Physics-Infused Neural MPC of a DC-DC Boost Converter with Adaptive Transient Recovery and Enhanced Dynamic Stability

Abstract

DC-DC boost converters require advanced control to ensure efficiency and stability under varying loads. Traditional model predictive control (MPC) and data-driven neural network methods face challenges such as high complexity and limited physical constraint enforcement. This paper proposes a hybrid physics-informed neural network (PINN) combined with finite control set MPC (FCS-MPC) for boost converters. The PINN embeds physical laws into neural training, providing accurate state predictions, while FCS-MPC ensures constraint satisfaction and multi-objective optimization. The method features adaptive transient recovery, explicit duty-ratio control, and enhanced dynamic stability. Experimental results on a commercial boost module demonstrate improved transient response, reduced voltage ripple, and robust operation across conduction modes. The proposed framework offers a computationally efficient, physically consistent solution for real-time control in power electronics.
Paper Structure (18 sections, 29 equations, 3 figures, 1 table)

This paper contains 18 sections, 29 equations, 3 figures, 1 table.

Table of Contents

  1. Introduction
  2. Preliminaries
  3. Finite Control Set Model Predictive Control (FCS-MPC)
  4. Prediction Model
  5. Cost Function Minimization
  6. Multi-Objective Terms & Weighting Factor Design
  7. Physics-Informed Neural Network (PINN)
  8. The Building Blocks of PINNs
  9. Ordinary Differential Equations (ODEs)
  10. Partial Differential Equations (PDEs)
  11. Loss Function
  12. Methodologies: Hybrid PINN+FCS-MPC
  13. data-driven Step
  14. phys-driven Step
  15. FCS-MPC Layer
  16. ...and 3 more sections

Figures (3)

  • Figure 1: Power conversion layout for variable-speed WECS. (a) System architecture, and (b) Key waveforms of the boost converter under CCM.
  • Figure 2: Conceptual illustration of the high-fidelity PINN+FCS-MPC framework. (a) Receding horizon policy, (b) Governing principle of the conventional ANN+FCS-MPC framework, (c) Governing principle of the proposed PINN+FCS-MPC framework, (d) FCS-MPC control structure in MATLAB/Simscape (left-panel) and MPC Flowchart (right-panel), (e) Hierarchical PINN+FCS-MPC offline deployment pipeline (left-panel) and PINN results (right-panel).
  • Figure 3: Experimental setup and performance evaluation of the CUT with PINN+FCS-MPC assisted control. (a) Prototype testbed ($\textcircled{1}$ DMM, $\textcircled{2}$ O-scope, $\textcircled{3}$ PSU, $\textcircled{4}$ DC EL, $\textcircled{5}$ TI-PMLKBOOSTEVM, $\textcircled{6}$ TI-$\mu$C, and $\textcircled{7}$ host PC), (b) Measured $V_{\mathrm{sw}}$ under different duty ratios, (c) Simulated steady-state operation under CCM$\dashrightarrow$BCM$\dashrightarrow$DCM, (d)–(f) Experimental steady-state waveforms under constant-current loads of 0.6 A, 0.4 A, and 0.3 A, (g)–(j) Output voltage TR to load current steps from 1 A to 0.2 A and from 0.2 A to 1 A, w/o and w/ PINN+FCS-MPC assist, (k)–(l) Load TR in CR mode: $200 \ \Omega \dashrightarrow 500 \ \Omega$.