Real-Time Digital Twin for Multiparameter Estimation and Health Assessment in a Two-Stage Grid-Connected Power Electronic Converter

This article delineates a novel digital twin (DT)centric methodology for parameter estimation and component health assessment within a two-stage single-phase grid-connected solar photovoltaic (SPV) inverter system. The DT, designed as a mathematical model interacting with the physical system (PS), is carefully formulated using discrete-time domain equations using the backward Euler approach. Sensor data collected from PS acts as a medium, enabling real-time interaction between digital and physical elements. The objective function defines the integration of PS-sampled data with the DT, describes the precise representation of the system. A novel advanced momentum search algorithm (AMSA), a physics-informed meta-heuristic optimization technique, is introduced to optimize essential parameters, such as capacitor and inductor losses, exhibiting enhanced convergence and accuracy compared to traditional methods. For component health assessment, experimental tests across various operating situations are performed in real time. The DT for the experimental prototype is validated using FPGA-based real-time OPAL-RT and yields an error margin less than 1.85%, thus highlighting the importance of DT in advancing real-time monitoring, health assessment, and parameter optimization. © 2025 IEEE. All rights reserved,