Traditional methods such as Steinmetz's equation (SE) and its improved variant (iGSE) have demonstrated limited precision in estimating power loss for magnetic materials. The introduction of Neural Network technology for assessing magnetic component power loss has significantly enhanced accuracy. Yet, an efficient method to incorporate detailed flux density information—which critically impacts accuracy—remains elusive. Our study introduces an innovative approach that merges Fast Fourier Transform (FFT) with a Feedforward Neural Network (FNN), aiming to overcome this challenge. To optimize the model further and strike a refined balance between complexity and accuracy, Multi-Objective Optimization (MOO) is employed to identify the ideal combination of hyperparameters, such as layer count, neuron number, activation functions, optimizers, and batch size. This optimized Neural Network outperforms traditionally intuitive models in both accuracy and size. Leveraging the optimized base model for known materials, transfer learning is applied to new materials with limited data, effectively addressing data scarcity. The proposed approach substantially enhances model training efficiency, achieves remarkable accuracy, and sets an example for Artificial Intelligence applications in loss and electrical characteristic predictions with challenges of model size, accuracy goals, and limited data.@en

Resonant converters are popular in power electronics due to their soft-switching capabilities, which enhance efficiency and prolong component lifetime. Three- phase resonant converters are particularly noteworthy for their higher power density and reduced ripple, making them ideal for demanding applications. A critical aspect of optimizing three-phase LLC resonant converters is the design of a transformer with adequate leakage inductance required for the resonance circuit. This paper compares two distinct transformer designs for such converters: a five-limb shell-type transformer and a symmetrical triangular transformer. Both designs are evaluated in terms of their performance, efficiency, and suitability for integration into the converter architecture. A detailed design procedure using Finite Element Method (FEM) analysis is presented to guide the development of these transformers. The practicality of this approach and its effectiveness are demonstrated through the implementation of a 3.4 kV to 60 V, 50 kVA prototype. This work provides a comparative analysis of transformer designs and introduces a validated methodology for improving the performance of three-phase LLC resonant converters through optimized transformer design.

@en

In the production of green hydrogen, electrolyzers draw power from renewable energy sources. In this paper, the design of Solid State Transformer (SST) for large-scale H ₂ electrolyzers is benchmarked. The three most promising topologies are chosen for design and comparison, including Modular Multi-level Converter (MMC) based SST, Modular Multi-level Resonant (MMR) based SST, and Input-Series-Output-Parallel (ISOP) based SST. The distance between converter towers for insulation and maintenance, the insulation system of the transformer, and the cooling system are designed with practical considerations in order to have an accurate estimation of the volume and weight of the SST. Losses in the switches are calculated based on equations, and losses in passive components are calculated based on FEM simulation. The operating frequency for each topology is optimized to minimize loss, weight, and volume. The best of each topology is then compared with each other to identify the most suitable one for large-scale H ₂ electrolyzers.

@en

This paper summarizes the main results and contributions of the MagNet Challenge 2023, an open-source research initiative for data-driven modeling of power magnetic materials. The MagNet Challenge has (1) advanced the state-of-the-art in power magnetics modeling; (2) set up examples for fostering an open-source and transparent research community; (3) developed useful guidelines and practical rules for conducting data-driven research in power electronics; and (4) provided a fair performance benchmark leading to insights on the most promising future research directions. The competition yielded a collection of publicly disclosed software algorithms and tools designed to capture the distinct loss characteristics of power magnetic materials, which are mostly open-sourced. We have attempted to bridge power electronics domain knowledge with state-of-the-art advancements in artificial intelligence, machine learning, pattern recognition, and signal processing. The MagNet Challenge has greatly improved the accuracy and reduced the size of data-driven power magnetic material models. The models and tools created for various materials were meticulously documented and shared within the broader power electronics community.

@en

The State of Health (SOH) is a crucial component of battery management systems (BMSs), offering important health information and protection against unsafe usage. In this paper, an accurate model for SOH estimation of Li-ion batteries was developed, which is uniquely characterized by using only the imaginary part of impedance at a specific frequency for precise SOH estimation. Through the identification of the relationship between impedance at a specific frequency and capacity degradation using correlation coefficients, the feature data most closely related to battery aging was selected. Next, the battery aging modeling and SOH estimation were validated on nine batteries across three different temperatures using a Feed-forward Neural Network (FNN). The validation results indicated that the proposed method has a high estimation accuracy, achieving a Mean Absolute Percentage Error (MAPE) of merely 2.05% throughout the entire lifecycle of the battery 45C02 during tests at a temperature of 45°C.

@en

Green hydrogen production uses renewable energies to energise the electrolysers for hydrogen production. The present paper compares possible solutions and configurations of a medium-frequency transformer (MFT) as part of a solid-state transformer (SST) in green hydrogen production applications. The single-phase and three-phase MFTs are compared and it is shown that a Yd three-phase MFT is the optimum choice for applications that require high power delivery and step-down of the voltage. A summary of previous works about MFT is also provided. Three-phase SST based on modular multilevel converters (MMC) is then described and various cases are investigated to obtain the optimum operational frequency. A 25 MVA, 400 Hz, 25.4 kV / 560V oil-immersed MFT design is presented and is shown that the proposed 400 Hz transformer saves 69% of the active parts' weight compared to a conventional line-frequency transformer (LFT).@en

For electrolyzer applications, traditional solutions using line frequency transformers plus rectifiers are bulky, heavy and have low controllability. The Solid State Transformer (SST) could be a promising solution to solve the mentioned issues. This paper compares the semiconductor ratings and capacitance of five different Modular Multilevel Converter (MMC) based Solid State Transformer (SST) topologies. The results show that the DRU (Diode Rectifier Unit)-MMC based topologies have the lowest semiconductor ratings and capacitance. Because of the unidirectional power flow requirement, the source side MMC of Back-to-Back (BtB) MMC based SST could be replaced by DRU, thus the cost is drastically saved. Another interesting finding is that the DRU-MMC energy ripple is much lower than half of the energy ripple in BtB MMC, which is different from HVDC MMC.

@en