Dynamic Charging Method for Fast Charging of Lithium-Ion Batteries and State of Health Estimation

Lithium-ion batteries (LIBs) are widely used in electric vehicles (EVs) due to their excellent characteristics, including high energy density, energy efficiency, long cycle life, and power stability. However, compared to internal combustion engine vehicles(ICE Vehicles), the long charging time remains a major limitation of EVs. To resolve this, high-current charging has been explored. Nevertheless, applying high current can lead to reduced energy efficiency, capacity fade, and power degradation, thereby revealing the limitations of simply increasing the charging rate. To overcome these issues, extensive research is being conducted not only on battery materials but also on charging strategies. This study explores the potential of suppressing battery degradation and extending battery life by employing a dynamic charging technique that introduces brief discharge patterns during the charging process. This novel charging method, which involves applying discharge pulses during charging, is experimentally compared with conventional constant current (CC) and multistage constant current charging methods to evaluate its effectiveness in reducing degradation. Previous studies have reported that under discharge conditions, dynamic cycling patterns that mimic real driving conditions—such as regenerative braking and rapid acceleration/deceleration—can more effectively suppress battery degradation than standard CC-based discharge cycles. This study extends that concept to the charging domain, hypothesizing that applying similar dynamic patterns during charging may also contribute to degradation mitigation. Commercial 18650 cylindrical lithium-ion cells were used for the experiments. Dynamic charging patterns with short discharge pulses were compared with CC charging across different state-of-charge (SOC) ranges (low, medium, and high). Throughout the charge-discharge cycles, degradation indicators such as capacity fade and increases in direct current internal resistance (DCIR) were periodically monitored using Reference Performance Tests (RPT) and Hybrid Pulse Power Characterization (HPPC) protocols. The results demonstrated that dynamic charging patterns had a more significant degradation suppression effect in certain SOC ranges compared to others. Additionally, an artificial intelligence (AI) model was applied to estimate state of health (SOH) and DCIR based on the voltage and current response data collected during the pulse tests. This approach presents a novel method for diagnosing battery condition solely using pulse responses generated during the charging process, without the need for additional testing equipment or disassembling the cells. By utilizing the brief discharge pulses that occur repeatedly in every charging cycle, this method enables continuous monitoring of battery degradation status, offering improved efficiency and practicality over traditional diagnostic techniques. In conclusion, this study proposes a novel charging method capable of mitigating battery degradation. Beyond degradation reduction through dynamic charging patterns, the study also presents an innovative approach wherein the charging process itself functions as a diagnostic tool. This contributes meaningfully to the fields of battery health diagnosis and life prediction. Furthermore, by optimizing SOC ranges and implementing data-driven charge control, the proposed method is expected to improve battery life and contribute to the development of lightweight diagnostic algorithms suitable for integration into future battery management systems (BMS).