An intelligent charging control algorithm for improving the efficiency and safety of new energy vehicle batteries

Efficient and safe battery charging is a critical factor influencing the operational performance, energy consumption, and service reliability of new energy vehicles (NEVs). With the rapid growth of electric mobility, improving charging efficiency while limiting battery degradation has become an important challenge for transportation energy systems. This study proposes an intelligent charging control framework that optimizes NEV battery charging efficiency under dynamic operating conditions. An adaptive deep learning–based algorithm combining a stacked long short-term memory (LSTM) network with an Adaptive Drosophila Food Search (ADFS) optimization strategy is developed to regulate charging current and voltage in real time. The proposed method incorporates multiple operational constraints, including state of charge, state of health, battery temperature, and terminal voltage, to balance fast charging performance, safety requirements, and long-term battery durability. Kalman filter–based preprocessing is applied to enhance the robustness of sensor data used for control decisions. Experimental results using benchmark battery charging datasets show that the proposed algorithm achieves higher prediction accuracy and improved charging optimization compared with existing methods, with an R² value of 0.95 and lower error indicators (MAE = 0.30, RMSE = 0.45). The findings indicate that the proposed approach can effectively reduce thermal stress, improve charging efficiency, and extend battery service life. The study provides a practical algorithmic solution for intelligent battery charging control in future NEV transportation and energy management systems.