Self-learning low-pass filter for energy management of hybrid energy storage electric vehicles

• A Self-Learning Low-Pass Filter (SL-LPF) EMS is proposed to improve power allocation in HESS electric vehicles. • The SL-LPF addresses key limitations of conventional LPF-based EMS, including phase shift, SoC constraints, cut-off frequency tuning, and adaptability. • PSO-based online learning enables SL-LPF to operate without prior track data. • SL-LPF achieves a 20.072% range improvement over battery-only EVs and performs comparably to data-intensive ILPF. • Validated across multiple drive cycles, SL-LPF outperforms conventional LPF and FLC LPF, demonstrating robust and adaptable EMS performance. Electric vehicles (EVs) offer a sustainable solution to reduce emissions in the transportation sector. However, they face challenges, such as high costs and limited onboard energy capacity. To address these issues, hybrid energy storage systems (HESSs) that combine batteries with supercapacitors have been introduced. This study proposes a self-learning low-pass filter (SL-LPF) energy management strategy (EMS) for HESS in EVs. Unlike conventional LPF methods, which suffer from phase shift, lack of state-of-charge (SoC) control, and fixed cut-off frequencies, SL-LPF integrates an online particle swarm optimization (PSO) mechanism to dynamically adapt filter parameters in real time. This approach requires only basic system information, making it practical for real-world and retrofitted EV applications. Simulation results show that the SL-LPF achieves a performance comparable to that of the offline-optimized improved LPF (ILPF) while operating with limited data. Across multiple drive cycles, the SL-LPF consistently outperformed the conventional LPF and fuzzy logic controller LPF (FLC-LPF). In the Worldwide Harmonized Light Vehicles Test Cycle (WLTC) cycle, the SL-LPF improved the vehicle range by 20.07% compared to a battery-only EV, with only a 0.19% gap compared to the ILPF. These findings highlight the originality and practicality of the SL-LPF as a low-complexity, data-efficient EMS solution for HESS EVs.