Unified Online Estimation Method for SOC, SOH, and Power Capacity Considering Safety Boundary Consistency in Battery Management Systems

Under high C-rate and wide-temperature conditions, independently estimated SOC and SOH often diverge due to decoupled model dynamics, resulting in inaccurate power boundary calculations. This affects power limiting, thermal safety, and fast-charging strategies. To solve this, a unified online estimation framework is proposed for SOC, SOH, and power capacity under voltage and thermal constraints. It integrates a state-space model based on an equivalent circuit, combining SOC, polarization voltage, internal resistance, and capacity degradation, with temperature-dependent parameter evolution to capture coupling with aging. A dual extended Kalman filter enables collaborative SOC–SOH estimation, while lightweight machine learning modules correct internal resistance and polarization dynamics to reduce mismatch under extreme conditions. Physical constraint projections embed voltage, temperature, and power limits into the estimation loop, mitigating noise amplification and drift. Based on consistent estimates, the SOP boundary is computed online to support control decisions. Validation across six temperatures (−20 °C to 55 °C) and five C-rates (0.2C to 6C), using bench, HIL, and pack-level tests over 120+ hours, shows SOC RMSE <1.6%, SOH error <2.5%, and SOP hit rate >95% within 10 seconds. Under noise and parameter disturbances, error growth is reduced by ~25% versus baselines. These results confirm improved SOC–SOH consistency and boundary tracking, with computational cost suitable for embedded deployment.