A novel hybrid approach for efficient energy management in battery and supercapacitor based hybrid energy storage systems for electric vehicles

The research work proposes optimal energy management for batteries and Super-capacitor (SCAP) in Electric Vehicles (EVs) using a hybrid technique. The proposed hybrid technique is a combination of both the Enhanced Multi-Head Cross Attention based Bidirectional Long Short Term Memory (Bi-LSTM) Network (EMCABN) and Remora Optimization Algorithm (ROA), termed the EMCABN-ROA technique. This approach involves an EMCABN technique that was trained online using load power requirements to match hybrid energy storage dataset characteristics for low-frequency power requirement prediction for the battery. The proposed technique is implemented in the Matlab and compared with several other benchmarks, such as Proportional Integral Derivative (PID) and Fractional-Order Proportional-Integral-Derivative (FOPID). The efficiency of regenerative braking using the proposed technique is reported to decrease to 4.5%. Efficiencies for different SOC ranges are analyzed, with the proposed approach at 48% and 43.5%, underscoring the significance of the initial voltage of the SCAP in the utilization of braking energy.