Accurate state-of-charge (SoC) estimation is fundamental to the reliable operation of lithium-ion battery systems in electric vehicles, stationary storage, and renewable energy integration. Traditional optimization algorithms struggle to identify battery parameters precisely under nonlinear conditions across varying temperatures, limiting the effectiveness of battery management systems.
This research addresses that challenge by introducing an Adaptive Honey Badger Optimization algorithm that combines sine-chaotic initialization with adaptive exploration and exploitation mechanisms. The approach dynamically balances the search between global scanning and local refinement, avoiding the local optima traps that plague conventional methods like Particle Swarm Optimization.
When applied to equivalent circuit models of lithium-ion cells, the AHBO method achieved terminal-voltage estimation errors as low as 7.85–16.64 millivolts across temperatures from 45°C to −15°C, outperforming standard Honey Badger and PSO approaches. More significantly, the algorithm reduced computational convergence time by approximately 70–75%, a substantial gain for real-time battery management applications.
Following parameter identification, the researchers compared two state-of-charge estimation techniques: the Extended Kalman Filter (EKF) and the Gated Recurrent Unit (GRU), a machine-learning approach. The EKF produced physically consistent results with SoC estimation errors of 0.87% at elevated temperature, 1.18% at moderate conditions, and 1.48% at low temperature. The GRU showed superior accuracy at intermediate temperatures but greater variability across the thermal range.
These findings have practical implications for battery management system designers. The AHBO algorithm enables faster, more accurate parameter estimation, improving the reliability of SoC prediction across real-world operating conditions. For battery-dependent sectors—from electric mobility to grid stability—reducing estimation uncertainty translates directly to improved performance, longevity, and safety margins.
