Conference Agenda

This paper extends a previously developed statistical methodology that uses Pearson's Random Walk (PRW) to model common mode (CM) current in multiple power electronic (PE) converter configurations.

The proposed methodology investigates emission reduction with respect to a single harmonic of the switching frequency in power electronic converters, specifically examining how additional converters affect the emissions of a single converter. Correlation analysis between fundamental, third, fifth, and seventh harmonics is performed to determine potential interrelationships.

The primary objective is to achieve electromagnetic emission reduction across the entire frequency spectrum rather than focusing on individual harmonics. Initial findings demonstrate that a worst-case relationship can be established between contributions at the fundamental frequency and contributions at higher-order harmonics. These results can potentially enable control schemes that reduce overall electromagnetic interference (EMI) in multi-converter setups.

This paper presents an extension of a Pearson's Random Walk (PRW) discussed in our previous papers as a modelling tool for predicting the common mode (CM) current in multi-converter setups, based on the CM current of a single converter.

Our earlier work was limited by the assumption that CM current could only be modeled as a sinusoidal damped oscillation. Here, we introduce a more general approach where the PRW modelling can be applied to any periodic signal with an existing Fourier Transform. Beyond theoretical discussion, we validate our extended methodology through multiple test cases using signals commonly found in power electronic (PE), electromagnetic compatibility (EMC), and other electrical circuit applications -- including damped sinusoids, exponential decays, chirps, Lorentzian pulses, half-sine pulses, Gaussian pulses, and sawtooth waves.

Our test results confirm our theoretical derivations, demonstrating that the Pearson's Random Walk (PRW) can be effectively extended to model diverse classes of periodic signals. We also address limitations associated with test signal post-processing.

Electromagnetic (EM) field simulation is essential to aircraft design for high intensity radio frequency (HIRF) qualification under RTCA DO-160. While a number of numerical full wave simulation tools have been shown capable of predicting HIRF test environments, model building is complex and solve times are typically long and/or expensive. Newer, stochastic power flow (SPF) methods are a mesh-free, wave physics alternative which greatly simplifies model building and solves 1000x faster than numerical codes. SPF formulations developed by the authors have recently been implemented in a new class of EM cable/field modeling software application - Stochastica. This paper documents a rigorous quantitative benchmark of the fast SPF simulation software, directly comparing with finite difference time domain (FDTD) numerical model results for a HIRF study of the Airbus A400M engine nacelle and wiring harnesses. The benchmark also demonstrates how SPF modeling statistically encloses the high sensitivity of numerical models to small perturbations in uncertain model parameters.

An electrical inverter-driven air propulsion system can produce unintended electromagnetic interference that disturbs the components or (sub-)systems in its vicinity. A realistic electromagnetic compatibility scenario for this aerospace use case starts with conducted common-mode emissions in the air propulsion system and its connected cable trees leading to radiated emissions that propagate within the fuselage. These, in turn, couple into a neighboring CAN bus on a (sub-)system's PCB and lead to differential-mode signals that potentially interfere with the CAN communication. The work presented here concerns the emulation of genuine power electronics driven radiated emissions inside a GTEM cell, exposing a CAN (sub-)system and characterizing the coupling into the CAN bus in detail. It will be shown, that based on expected conducted common-mode emissions in an electrical inverter-driven air propulsion system and typical distances between the cable tree and a victim CAN bus, sufficiently high differential-mode interference can be induced to affect the CAN communication.