Accurate extraction of equivalent circuit model (ECM) parameters is essential for aging-aware battery management in lithium-ion batteries. Electrochemical impedance spectroscopy (EIS) offers detailed insight into ohmic resistance, charge-transfer kinetics, and diffusion processes, but its onboard application is constrained by unstable measurement conditions. This work introduces a diagnostic window at 100% SOC immediately after the constant-voltage (CV) phase, where interfacial stabilization and kinetic relaxation yield quasi-equilibrium suitable for reproducible impedance measurements. A short post-CV rest is included only as a verification step to confirm minimal voltage drift. Validation was performed on three cells representing pristine, moderately aged, and heavily aged states. Galvanostatic Intermittent Titration, conducted at a low C/25 rate, provided a laboratory benchmark, while EIS was carried out at 0%, 50%, and 100% SOC under controlled rests. Comparative analysis showed strong consistency in ohmic resistance across techniques, while EIS demonstrated superior resolution of charge-transfer and diffusion processes, particularly in aged cells, thereby making it suitable for real-time evaluation. These findings establish the CV-based diagnostic window as a reproducible, diagnostically rich, and onboard-compatible method for ECM parameter tracking and aging diagnostics.

This article presents a cloud-edge deployed, physicsguided bidirectional long short-term memory (Bi-LSTM)-based framework for real-time core temperature estimation of automotive lithium-ion battery (LIB), enabling enhanced thermal safety and predictive control. Unlike existing approaches that rely on surface temperature sensors or offline models, the proposed framework integrates standard BMS signals (voltage, current, surface, and ambient temperature) with physics-guided feature engineering to capture electrothermal dynamics while maintaining low computational cost. The model, trained across a wide range of C-rates, dynamic drive cycles, and ambient conditions, achieves a mean absolute error (MAE) of 0.16 degrees C and an RMSE of 0.26 degrees C, outperforming comparable sequence-learning architectures. Real-time validation demonstrates accurate estimation across unseen cells, achieving 0.31 degrees C MAE and 0.40 degrees C RMSE at the module level. Integration with CAN-based closed-loop control improves the thermal response by at least 2 min compared to state-of-the-art surface-temperature-based strategies. This improvement provides a critical safety margin for preventing thermal runaway. The framework is deployed on both local and cloud servers, achieving latency as low as 30 ms (local) and 85 ms (cloud), with real-time visualization through InfluxDB-Grafana, enabling remote monitoring and long-term data storage.

This paper presents a deep learning–based framework for the automated classification of equivalent circuit models (ECMs) from electrochemical impedance spectroscopy (EIS) data for enhanced lithium-ion battery (LIB) diagnostics. EIS captures key internal electrochemical processes, including charge transfer, diffusion, and double-layer behavior, but conventional ECM identification remains labor-intensive and difficult to scale. To address this challenge, a one-dimensional convolutional neural network (1D-CNN) is developed to classify EIS spectra into four ECM categories using features derived from both real and synthetic datasets. The proposed network incorporates hierarchical convolutional layers, batch normalization, dropout, and global average pooling for robust feature learning and generalization. Experimental results demonstrate a high classification accuracy of 98.25% on the test dataset. The predicted ECMs show strong agreement with Nyquist plot characteristics, confirming the accuracy, robustness, and practical relevance of the proposed approach for intelligent battery management systems.

The increasing adoption of lithium-ion batteries (LIBs) in electric vehicles necessitates accurate and continuous temperature monitoring to ensure optimal performance, safety, and longevity. This paper presents a study using PhysicsInformed Neural Network (PINN)-based approach for realtime estimation of LIB surface temperature, integrated within the Battery Management System (BMS). A detailed simulation study is first conducted in MATLAB/Simulink by modeling the charge-discharge behavior of a battery pack in a parallel hybrid electric vehicle under the UDDS drive cycle. The drive cycle profile is implemented using a lookup table to generate realistic operational data. The generated dataset is subsequently used to train the PINN model for surface temperature prediction. The proposed approach achieves a prediction error of $0.36^{\circ} \mathrm{C}$ under simulated conditions. Furthermore, the model is validated using experimental data from a $\mathbf{1 4 S}$ lithium-ion battery pack subjected to the UDDS drive cycle, demonstrating an improved prediction accuracy of $0.12^{\circ} \mathrm{C}$. The results highlight the effectiveness of the proposed method for accurate and real-time thermal estimation in LIB-based systems.

Reliable estimation of internal battery temperature is essential for safe operation and fast charging of electric vehicles. However, core temperature cannot be measured directly in practical BMS, and data-driven estimators may lose accuracy under operating-condition shifts, measurement noise, and uncertain feedback. This paper proposes a diagnostic-driven adaptive learning framework for robust battery thermal estimation and reinforcement-learning-based charging control. The proposed approach analyzes learning signals using interpretable diagnostics that capture systematic bias, stochastic noise, and update alignment, and uses these diagnostics to regulate parameter updates during training, adaptive online learning, and control. A Bi-LSTM model is used as the backbone for battery core temperature estimation and is enhanced using the proposed optimization strategy. Experiments using measured battery data across multiple operating regimes demonstrate improved robustness and generalization. Under unseen C-rate conditions, the proposed method reduces mean absolute error from 0.391 °C to 0.295 °C and eliminates temperature underestimation events that may lead to safety risks. For cross-condition adaptation, the proposed meta-learning prior reaches the performance of conventional optimization methods trained for fifty adaptation steps while requiring only ten steps. In reinforcement-learning-based charging control with reward noise of standard deviation 0.25, the proposed method reduces return variability from 91.48 to 49.56. These results demonstrate improved reliability of battery thermal estimation and more stable learning-based charging control.

Solid-state transformers (SSTs) represent a significant advancement in modern power systems, offering improved efficiency, adaptability, and compatibility with renewable energy sources, smart grids, and electric vehicle (EV) infrastructure. However, their unique design, including high-frequency operation and advanced semiconductor components, poses specific challenges that require effective protection mechanisms. This article investigates modern protection strategies compatible with SSTs, focusing on overvoltage and overcurrent protection, thermal management, and short circuit fault detection. The discussion also explores the integration of smart technologies, advancements in materials, and enhanced thermal performance as critical areas for future development. To support the discussion, a case study is presented, detailing the implementation of a solid-state protection system integrated with an SST, developed by the authors. Advanced protection strategies are highlighted with key challenges to enhance the reliability of SSTs in dynamic power systems. Additionally, the article explores future directions for improving SST protection mechanisms.

Electrochemical impedance spectroscopy (EIS) is a powerful diagnostic technique for lithium-ion batteries, capable of probing and interpreting the internal phenomena such as ohmic resistance, charge-transfer kinetics, and diffusion processes. Recent developments in dynamic EIS (DEIS) have highlighted its potential for real-time monitoring of degradation and safety-critical events. However, these demonstrations rely on laboratory potentiostats or three-electrode research cells, which are impractical for integration into onboard battery management systems (BMS). This work investigates the feasibility of using a compact, low-power impedance monitoring integrated circuit for embedded EIS measurements under static conditions. The Analog Devices AD5941 evaluation platform was configured in a two-electrode, four-point Kelvin setup with bias control, sinusoidal excitations and a frequency range extending into the kilohertz domain. Impedance data were collected via UART for post-processing. Nyquist and Bode responses show good agreement with reference instrument data in the mid- and high-frequency ranges, demonstrating the suitability of chip-level hardware as a foundation for future EIS-enabled onboard diagnostics.

This paper examines optimal design principles and practices essential for the printed circuit board (PCB) layout of switching regulators with fast dv/dt and di/dt edge rates to achieve electromagnetic interference (EMI) compliance standards. Reflections, crosstalk, and transmission line management techniques along with power delivery network/system (PDN/PDS) design are implemented for enhanced signal and power integrity. 4-layer board designed with embedded interplane capacitance, controlled impedance traces, proper signal termination and crosstalk management exhibits lowest noise emissions. The experimental results show good consistency with electromagnetic field theory and simulations. The paper also outlines effective design rules and practices essential for optimizing EMI/EMC performance and power integrity during the layout process of switching regulators in PCBs. By employing robust transmission line design techniques and power delivery system (PDS) design principles, a reliable PCB design for the DC/DC converter is achieved, ensuring compliance with EMI standards, and maintaining power integrity.

The rapid advancement of battery management systems (BMS) in automotive applications demands real-time, automated data acquisition, and visualization architectures capable of handling complex battery dynamics. This article introduces a robust, scalable cloud-based architecture that seamlessly integrates with the physical on-board BMS for enhanced monitoring, predictive analytics, and long-term data storage. The system uses automotive-grade hardware, including an NXP BMS and STM32 microcontroller and an efficient Python-based CAN data decoding algorithm to enable accurate real-time monitoring and visualization via Grafana. Comprehensive experiments reveal the system's efficiency in tracking critical parameters like cell voltage, temperature, and balancing voltage, ensuring proactive detection of weak and faulty cells, thereby improving battery safety. Key contributions include high-resolution, precision real-time battery data sampling; efficient CAN data decoding; data safety and security; identification of weak cells; and analysis of how data sampling rates and cloud server locations impact communication latency, memory usage, and computational power. Understanding these factors is crucial for the scalability of cloud-based BMS in automotive applications. The proposed architecture will aid in the practical implementation of cloud-enhanced BMS and digital twin-based BMS. It will also benefit second-life applications of retired automotive batteries due to long-term historical data storage.

This paper proposes an enhanced control strategy for a three-phase, two-level power factor correction (PFC) rectifier cascaded with a dual-active-bridge (DAB) DC-DC stage for Level-3 electric vehicle (EV) chargers. High-power fast chargers require a near-unity power factor and compliance with IEEE-519 harmonic standards. The proposed system employs silicon-carbide (SiC) MOSFETs and a voltage-oriented control framework with a Type-II compensator for outer DC-link regulation and synchronous dq-frame current control. Space Vector PWM (SVPWM) is implemented using a projection-based algorithm that eliminates trigonometric computations and sector detection by directly projecting the reference vector with symmetric zero-vector timing, resulting in a compact, real-time-friendly solution. The stabilized DC-link voltage drives the DAB converter under single-phase shift modulation. Real-time simulation studies on OPAL-RT demonstrate fast transient response, IEEE-519 compliant current quality, and near-unity power factor. The proposed controller outperforms conventional PI-based SRF methods and maintains robust performance even under non-ideal grid conditions, confirming its suitability for high-power EV charging.

Characterisation of internal electrochemical behaviour of a Lithium-ion battery (LIB) is crucial to assess performance, safety and lifespan. Combining different characterisation techniques complements the understanding and characterizes battery internal processes. In this study, the correlation between a time-domain characterization technique, galvanostatic intermittent titration technique (GITT), and a frequency domain, electrochemical impedance spectroscopy (EIS) technique for parameterization is studied. Both techniques provide insights into the kinetic and diffusion behaviour of LIB. However, both techniques characterize the battery considering different approaches. GITT measures overall polarization and diffusion coefficients using voltage transient response and relaxation time through RC components. Whereas EIS deconvolutes the internal behaviour of LIB into bulk resistance, charge transfer resistance and diffusion resistance. The experimental study reveals a strong correlation between ohmic resistance values derived from both techniques. Comparison studies between the two techniques provide insight into charge transfer and diffusion processes. This integrated approach advances LIB parameterization, aiding in the development of robust battery management systems and diagnostics for improved reliability and efficiency.

Solid-state batteries (SSBs) are at the forefront of next-generation energy storage technologies, promising enhanced safety, energy density, and longevity compared to traditional lithium-ion batteries. However, managing the thermal behavior of SSBs during charging is critical to ensuring their safe and reliable operation. This paper presents an in-depth analysis of the thermal characteristics of SSBs under various charging conditions, based on laboratory experiments conducted in controlled environments. The study specifically evaluates the thermal performance of SSBs in comparison to traditional lithium nickel cobalt aluminum oxide (NCA) batteries, with an emphasis on their potential application in e-mobility. The insights gained from this research are expected to inform the development of advanced thermal management strategies, contributing to the design of safer and more efficient solid-state battery technologies for e-mobility applications.

Health-conscious battery management systems (BMSs) that rely on surface temperature measurements are insufficient for managing automotive lithium-ion batteries (LIBs). Experimental studies have shown temperature differences of up to 10 degrees C between surface and core of cylindrical LIBs. BMSs that consider only surface temperature overlook critical thermal information. The missing monitoring can delay detecting thermal events within the cell, accelerating battery degradation and increasing the risk of thermal runaway. This article introduces two deep learning algorithms to address this: Kolmogorov-Arnold network (KAN) and interconnected long short-term memory (LSTM) network. Both approaches estimate the core temperature of LIBs without requiring surface temperature feedback to the neural network. Experimental validation revealed a core temperature mean absolute error (MAE) of 0.55 degrees C with a computational cost of 2.9-3.2 ms for KAN. The proposed interconnected LSTM reached a MAE of 0.80 degrees C. The performance of the two core temperature estimation techniques was further evaluated under dynamic loading profile using urban dynamometer driving schedule (UDDS) drive cycle. The KAN method achieved a MAE of 0.325 degrees C, demonstrating its adaptability to dynamic operating conditions. The two proposed methods, primarily KAN, are both adaptive and computationally efficient, making them suitable for integrating onboard BMS and cloud-enabled digital-twin-based BMS systems.

This paper presents a digitally controlled Inductive power transfer (IPT) system for charging Electric scooters (E-scooters), utilising an adaptive hybrid compensation network design. This system ensures constant current (CC) and constant voltage (CV) charging, with misalignment. It leverages the advantages of double-sided inductor-capacitor-capacitor (DS-LCC) and LCC-series (LCC-S) topologies, which respectively provide CC and CV output. A Type-II digital anti-windup PI control method with a selectable compensation network is introduced, allowing both CC and CV modes to operate with zero voltage switching (ZVS) under misalignment conditions. This approach reduces system losses and enhances the stability and efficiency of the IPT system. A 270W/85-kHz IPT-based E-scooter charger was designed and simulated in the MATLAB/Simulink environment, and an experimental prototype was developed to evaluate the performance of the proposed charger as per SAE J2954 standards. Testing under three different coupling conditions perfect alignment, 5 cm misalignment, and 10 cm misalignment demonstrated that the output parameters remained constant across all conditions, confirming the effectiveness of the proposed control technique.

A cloud-enhanced, real-time control architecture for reconfigurable battery systems (RBS) is introduced. The proposed system leverages cloud and fog computing to optimize state estimation and control, focusing on state-of-charge. The architecture is tested with a hybrid cascaded multilevel converter and evaluated concerning latency and bandwidth in hybrid and centralized compute settings. The results indicate that cloudbased control offers promising performance, particularly in latency and efficient bandwidth use. The shortest mean and median request time could be archived with the cloud, centralized, classic control approach with 71.3 ms and 67.7 ms. Furthermore, the data compression technique could reduce the amount of data transferred by 77.1% and 64.7% for the conventional battery pack and the RBS, respectively. This research closes a gap in cloud-based BMS by comparing classic and AI-based control in distributed cloud and fog computing setups.

This paper presents a dual-stage hybrid battery balancing control strategy designed for lithium-ion battery packs in electric vehicles (EVs). The proposed system integrates thermal management to minimize temperature rise while ensuring effective voltage balancing. An intelligent control algorithm dynamically adjusts the balancing currents in real time based on individual cell voltage and temperature, promoting battery longevity. The control architecture combines active cell-to-pack (C2P) balancing which is engaged during significant state-of-charge (SOC) imbalances along with passive balancing for finer adjustments, simultaneously operating within the safe voltage and thermal limits. This system is intended to complement existing Battery Management Systems (BMS), significantly accelerating the balancing process in second-life applications.

The rapid adoption of electric vehicles (EVs) and the growing need for high-power EV fast charging infrastructure have led to the widespread use of advanced power electronic topologies, notably the Dual Active Bridge (DAB) converter. While DAB-based DC fast chargers offer efficient bidirectional power flow and galvanic isolation, their integration into cyber-physical systems introduces new cybersecurity vulnerabilities. Existing research primarily focuses on communication protocols (e.g., OCPP, ISO 15118), backend networks, or electric vehicle supply equipment (EVSE) firmware security, often overlooking the converter level where actual power processing occurs as a critical attack surface. This study investigates recent literature on cyber threats specific to DAB-based fast charging systems and reviews resilience strategies that enhance system robustness. Emphasis is placed on the unique attack surfaces of the DAB architecture, potential impacts on the control loop and power flow, and the integration of real-time monitoring, secure communication protocols, and anomaly detection systems. The objective is to provide the EV charging and smart grid research communities with an understanding of emerging threats at the converter and its controller level and to propose recommendations for designing cyber-resilient EV fast charging systems. The paper concludes by outlining key research gaps and offering directions for future work to support the development of secure and reliable DAB-based EV fast charging infrastructure.

Solid-state batteries (SSBs) are emerging as a promising alternative to conventional lithium-ion batteries due to their superior safety, energy density, and lifespan. However, understanding their thermal behavior under dynamic operating conditions is crucial for ensuring safety and performance, especially in e-mobility applications. This study presents a comparative thermal analysis of SSB and lithium nickel cobalt aluminum oxide (NCA) 21700 cells during charging under various ambient temperatures (0 degrees C, 25 degrees C, and 40 degrees C). Key metrics such as temperature gradients (Delta T/Delta t) and differential temperature rise (Delta T) are evaluated to identify critical thermal behaviors. The results reveal that SSBs exhibit significantly higher Delta T and Delta T/Delta t. While battery management systems (BMS) typically regulate absolute temperature rise (Delta T), this study highlights the importance of monitoring Delta T/Delta t as a critical parameter for mitigating accelerated degradation and preventing thermal runaway. The findings contribute valuable insights toward developing robust thermal management strategies for next-generation battery systems.