In this article, a secure, sustainable, and fast charging strategy based on an advanced electro-thermal-degradation (ETD) model is proposed for lithium-ion batteries. In this ETD model, temperature, state-of-charge, and state-of-health are strongly coupled. It combines the simplicity of the equivalent circuit model and the high accuracy of the random forest (RF) algorithm. The RF algorithm is used to address uncertainties, such as measurement errors, parameter drifts, and parasitic effects, all of which may otherwise compromise model accuracy. The degradation model is derived from the Arrhenius equation and enhanced through data fitting techniques. To minimize charging time while respecting degradation constraints, we employ a data-driven Bayesian optimization (BO) approach with a constraint-based expectation improvement acquisition function. The proposed control strategy is implemented in a dual-loop structure, with an outerloop temperature control to ensure thermal safety and an innerloop current control to achieve sustainable and fast charging based on the BO. Experimental results demonstrate the effectiveness of the proposed strategy, showing that the ETD model achieves accurate temperature estimation across a wide range of operating conditions.

Nonuniform temperatures in lithium-ion battery modules, caused by manufacturing inconsistencies, vibrations, and unequal line resistances, lead to uneven current distribution and accelerated degradation of the battery. Existing thermal management methods face challenges in achieving real-time cell-level balancing due to limited intercell modeling, high computational cost, and lack of closed-loop control. This article proposes a model predictive temperature balancing control (MPTBC) strategy based on a scalable 2-D thermal network model (TNM) that captures intercell thermal coupling and enables real-time prediction with reduced computational cost. A physics-informed neural network (PINN) models the nonlinear internal resistance, with Bayesian optimization (BO) used to efficiently identify optimal parameters. The MPTBC is implemented on a four-module, high-power-density, single-input multioutput (SIMO) switched-capacitor (SC) converter. Experiments validate the TNM accuracy and demonstrate that MPTBC effectively minimizes cell-to-cell temperature differences.

Lithium-ion batteries (LIBs) dominate energy storage for electric vehicles (EVs) due to their high energy density, long cycle life, and low self-discharge. However, high costs, complex manufacturing, and the requirement for advanced battery management systems (BMSs) constrain their broader deployment. Therefore, extending the utility of LIBs through reuse is essential for economic and environmental sustainability. Retired EV batteries with 70–80% state-of-health (SOH) can be repurposed in battery energy storage systems (BESSs) to support power grids. Effective reuse depends on accurate and rapid assessment of SOH and state-of-safety (SOS), which relies on precise state-of-charge (SOC) detection, particularly for aged LIBs with elevated thermal and electrochemical risks. This review systematically surveys SOC, SOH, and SOS detection methods for second-life LIBs, covering model-based, data-driven, and hybrid approaches, and highlights strategies for a fast and reliable evaluation. It further examines power electronics topologies and control strategies for integrating second-life LIBs into power grids, focusing on safety, efficiency, and operational performance. Finally, it analyzes key factors within the closed-loop supply chain, particularly reverse logistics, and provides guidance on enhancing adoption and supporting the establishment of circular battery ecosystems. This review serves as a comprehensive resource for researchers, industry stakeholders, and policymakers aiming to optimize second-life utilization of traction LIBs.

The distributed convex optimization strategies are widely used for voltage regulation and current distribution among distributed energy storage systems (DESSs). Meanwhile, consensus-based secondary control is commonly employed but risks leaking initial state information when exchanged explicitly. To address this issue, a distributed state-decomposition convex optimization (DSDCO) approach is proposed for current distribution in DC networks, ensuring the protection of DESSs' initial states. In DSDCO, a state-decomposition strategy is implemented, where the state variable of each DESS is divided into two sub-state variables with random initial values. One sub-variable is dedicated to external consensus control, safeguarding the node's true initial state, while the other manages internal dynamics to achieve global consensus. The theoretical analysis confirms that DSDCO can achieve state consensus in finite time and effectively maintain privacy. Subsequently, the first state variable that incorporates consensus control is used for current allocation and loss optimization. Finally, the privacy-preserving effectiveness of DSDCO is validated through both simulation and experimental case studies.

The switched capacitor (SC) single-input-multioutput (SIMO) converter, known for its high conversion ratio, efficiency, and power density, has been widely adopted across various industrial sectors, including battery charging systems. However, the existing proportional-integral (PI) based master-slave control for the SC-SIMO converter may experience significant steady-state errors in the output currents when the converter's stability is strictly maintained over a wide operating range. To address this critical issue, this article proposes an artificial neural network (ANN) based virtual reference (VR) leader-follower control. The leader module is democratically chosen from all the modules to minimize the mean average error of the outputs. The leader module is regulated by a PI controller to follow the actual reference, while the follower modules are controlled by linear controllers to track the VRs generated by the ANN. Experimental results demonstrate the superior steady-state tracking performance of the proposed ANNVR control over conventional control under various operating conditions, without compromising the system's overall efficiency, dynamic performance, or incurring additional hardware costs.