Assessment of Parameter Identification Methods for Digital Twins of Two-Level Bidirectional Converters

Abstract

Monitoring and maintaining the health and performance of power electronic converters is of paramount importance in many applications. The use of digital twins for this purpose has received significant attention over the last few years. System parameter identification will be critical for the development of accurate and reliable digital twins. In this paper, the application of digital twin technology for monitoring a two-level bidirectional converter is explored, with a special focus on the selection of algorithms for parameter identification, and their implementation. While Particle Swarm Optimization (PSO) has gained increased attention in power converter applications recently, it does not necessarily produce the best results. Both simulation and experimental verification will be used to validate the findings. The final use of the proposed methods is onboard energy storage integration for railways. However, the conclusions reached in this paper can be extended to other applications as well as power converter topologies.