Random Walk Approach to predict electromagnetic emissions for multiple power electronic converters

Erjon, Ballukja

Random Walk Approach to predict electromagnetic emissions for multiple power electronic converters

Publication Date

December 31, 2025

Creator

Erjon, Ballukja

Abstract

This thesis addresses the increasing Electromagnetic Compatibility (EMC) challenges posed by multiple Power Electronic (PE) converters operating simultaneously within a network. As current EMC standards predominantly focus on single-device evaluation, a significant research gap exists in modelling and predicting aggregate electromagnetic interference from multiple converters. The research proposes a new application of Pearson’s Random Walk (PRW) theory to characterise Common Mode (CM) electromagnetic emissions in multi-converter configurations. The investigation demonstrates that Pearson’s Random Walk provides an effective statistical framework for modelling electromagnetic emissions from multiple PE converters, where traditional deterministic approaches have proven inadequate. The model is based on the assumption that the sole variable under control is the switching-on time of the converters. The model employs vectors that represent the phase of waveforms being produced by each converter, associating converter switch-on times with vector angles to predict aggregated electromagnetic interference. This approach was verified through both simulation studies of eight identical converters and experimental measurements with three DC/DC converters. Statistical verification through empirical and theoretical cumulative distribution function (cdf) confirmed the model’s validity regardless of harmonic number. Furthermore, the research presents the first explicit computation of the probability that electromagnetic interference is reduced in a multi-converter configuration compared to a single-converter arrangement. Results indicate that whilst electromagnetic interference reduction is possible, this probability diminishes with an increasing number of converters. The developed methodology offers manufacturers and network operators a robust framework for predicting worst-case electromagnetic emissions in multi-converter systems, thereby addressing requirements specified in current electromagnetic compatibility directives. This contribution advances the standardisation efforts of the IEC CISPR Working Group 4 concerning the impact of increased device quantities on electromagnetic compatibility.

Item Type

ethesis

Thesis Type

PhD

Supervisors