The surge in electric vehicle adoption has resulted in a significant rise in end-of-life batteries, which are unsuitable for demanding EV applications. Repurposing these batteries for secondary applications presents a promising avenue to tackle environmental and economic challenges associated with their disposal. The second-life battery (SLB) approach emerges as a mechanism to manage this massive amount of retired EV batteries. However, this approach poses significant challenges in determining and monitoring battery degradation and performance. After evaluating different scenarios for reusing or recycling retired EV batteries, this paper examines the main challenges associated with SLBs, including techno-economic aspects, uncertainty from first life, safety, characterization and screening, battery-management systems, and secondary applications. A comprehensive review of current state-of-the-art SLB research and implementations is provided, particularly emphasizing battery characterization and the requisite evaluation processes for SLB eligibility. This paper explores diverse measurement techniques for assessing SLB performance, evaluating them based on accuracy, complexity, and time consumption, which are essential for achieving cost-effective SLB applications. The overarching objective is to thoroughly understand the principal challenges associated with repurposing EV batteries and delineate the research imperatives necessary for their successful implementation and prolonged lifespan.

Li-ion batteries' complex and dynamic electrochemical behavior pose a significant challenge for their diagnostics and prognostics. Electrochemical Impedance Spectroscopy (EIS) is a valuable measurement technique that provides crucial insight into battery behavior. Transforming the EIS impedance data across various frequencies into equivalent electrical circuit elements tailored to the battery's physical and chemical characteristics is valuable for interpreting this data. Distribution of Relaxation Times (DRT) analysis is such an approach and results in a distribution of time constants that effectively characterize the RC networks of the battery's equivalent circuit model (ECM). Modeling the battery with detailed consideration of EIS data and minimizing the information that is lost in this modeling stage is the crucial advantage of the DRT technique. In this work, we present a fully automated procedure for modeling a substantial volume of EIS data using DRT. This involves the automatic selection of prominent peaks that persist throughout the entire battery aging process, effectively enhancing computational efficiency. It furthermore includes data pre-processing steps and frequency range selection, making DRT analysis more efficient and robust by employing the least squares fitting (LSQF) algorithm. The findings in applying our procedure to published EIS spectra of Li-ion battery cells indicate that the presented approach yields a high level of precision in tracking the ECM parameters to analyse the degradation of these cells. Further proof of the quality of our procedure lies in the consistent low dispersion of the various fits, which remain consistent across the entire dataset.

High-frequency core loss modeling plays a critical role in the magnetics design of power electronics. However, existing modeling tools fail to achieve both high speed and high precision. The conventional analytical approach enables fast estimations but performs poorly in accuracy. Magnetic loss models aided by loss maps feature high accuracy, but their model parameterization relies on large data. The emerging approach of artificial neural networks (ANNs) provides a promising alternative since it can achieve high speed and accuracy. However, conventional implementations of ANN require a large and accurate dataset for training, which is hard to achieve in magnetic loss modeling. To solve this problem, a knowledge-aware artificial neural network (KANN) is proposed that can achieve high accuracy with small training datasets. After introducing the principle of the proposed KANN, it is applied to high-frequency core loss modeling using the improved generalized Steinmetz equation as additional knowledge. To validate the performance of the proposed KANN-based design method for core loss modeling, it is applied to predict the losses of two ferrite cores in the frequency range of 50–450 kHz. The results show that the proposed method greatly outperforms present loss modeling approaches in accuracy and speed, requiring only a limited training dataset. An automatic loss modeling tool based on the new method is provided together with its open-source code.

In this paper, a reconfigurable isolated dc/dc converter is proposed suitable for a wide output voltage range of 180-1500V for efficient onshore charging in maritime applications. The proposed circuit concept can also fulfill the requirements of other heavy-duty battery charging applications, especially those operating within the shore environment-such as straddle carriers, forklifts, reach stackers, terminal tractors, etc. The circuit topology consists of two interleaved LLC resonant converters each connected to a three-winding transformer. Through the use of additional circuitry, the topology can be adapted to operate at peak efficiency in three output voltage ranges. Furthermore, the topology is able to alleviate the current and voltage stresses on the semiconductor devices in comparison to the conventionally employed LLC resonant converters. The operation of the circuit is explained and its steady-state model is developed. In order to validate the performance of the converter, an 11kW prototype is designed, tested, and analyzed. The experimental results attest that the proposed reconfigurable resonant converter (RRC) is able to achieve an extremely wide output voltage range while maintaining a high power transfer efficiency.

Due to their easy scalability to high power, simple control mechanism, and effective semiconductor and transformer utilization, H-bridge-derived isolated dc/dc converters have taken a central stage in various power electronics applications such as fast charging stations, dc microgrids, and renewable energy integration. Among the bidirectional variants, the dual active bridge (DAB) converter stands out as a common choice due to its direct power flow control mechanism, inherent soft switching realization, symmetric and modular structure, etc. However, the conventional version of the topology exhibits a significant reduction in efficiency when subjected to wide ranges in voltage gain and output power. This limitation can potentially hinder its application in public dc chargers, where different charging profiles and vehicle voltage classes are frequently expected. This article proposes a new versatile topology derived from the DAB converter that caters to the requirement of a wide output voltage range and improves the efficiency characteristics while mitigating the current and/or voltage stresses on the semiconductor devices. Furthermore, a suitable modulation strategy is developed that optimizes the soft switching action of the converter in accordance with the load demand. The converter operation is explained and validated through a steady-state model, simulations and an 11 kW experimental prototype.

Electrochemical impedance spectroscopy (EIS) holds great promise for assessing battery degradation. Nevertheless, many present methods for extracting battery aging features from EIS data are unsuitable for cells that have very different aging behaviour. Another challenge is that the time complexity of machine learning (ML) models based on full-frequency EIS significantly increases. To tackle these issues, this paper investigates two feature extraction methods together with suitable regression algorithms. Firstly, manual feature extraction methods are studied that involve feature extraction via feature correlation and consistency analysis, enhancing the robustness of battery capacity estimation within Gaussian process regression (GPR) and support vector machine (SVM) algorithms. On the other hand, automatic feature extraction methods extract latent features using convolutional neural networks (CNN) as inputs to the model. Best results are achieved with the CNN-based methods that achieve robust capacity estimation with 2.88 % accuracy for different battery cells at varying temperatures while remaining insensitive to EIS measurement uncertainty. This is a factor of 2 to 3 more accurate than existing published approaches on the same dataset. By combining the advantages of manual and automatic feature extraction approaches, a reduced EIS data set is used to train a lightweight CNN model, resulting in an inference time of less than 21.48 ms on STM32L476RG, whereas existing approaches take up to 50 ms for achieving a similar accuracy.