State-of-charge (SOC) estimation in batteries is a critical parameter for battery management systems (BMS), particularly as batteries age and their performance characteristics evolve. Accurate SOC estimation must account for capacity fade, resistance growth, and shifting voltage-SOC relationships over time. These aging effects introduce complexities that require advanced modeling and estimation techniques to maintain precision throughout a battery's lifecycle.

Capacity fade is one of the most significant aging effects impacting SOC estimation. As a battery cycles, active material degradation, lithium inventory loss, and electrode structural changes reduce the maximum available capacity. Traditional SOC estimation methods relying on coulomb counting become increasingly inaccurate if the assumed capacity does not reflect the actual degraded capacity. To compensate, modern SOC estimation algorithms incorporate capacity fade models that track the relationship between cycle count, operating conditions, and capacity loss. Empirical models based on accelerated aging data can predict capacity fade as a function of factors such as temperature, depth of discharge, and charge/discharge rates. These models enable real-time adjustments to the SOC calculation by scaling the estimated capacity to match the battery's current state-of-health (SOH).

Resistance growth is another critical aging effect that distorts SOC estimation. Increased internal resistance due to solid-electrolyte interphase (SEI) layer growth, contact resistance changes, and electrolyte decomposition alters the voltage response under load. Since many SOC estimation methods rely on voltage measurements, resistance growth introduces errors if not properly accounted for. Equivalent circuit models (ECMs) with adaptive parameters can track resistance changes over time, allowing the BMS to compensate for voltage deviations caused by aging. Online parameter identification techniques, such as recursive least squares (RLS) or Kalman filtering, continuously update the model's resistance values based on real-time operating data. This ensures that SOC estimation remains accurate despite increasing impedance.

The voltage-SOC relationship also evolves as a battery ages. Electrode polarization shifts, lithium plating, and active material loss can alter the open-circuit voltage (OCV) curve, which is often used as a reference for SOC estimation. Aging-aware SOC algorithms must adapt to these changes by periodically recalibrating the OCV-SOC relationship or using dynamic models that account for hysteresis and kinetic effects. Machine learning approaches, such as neural networks trained on aged battery data, can capture nonlinear voltage-SOC shifts and improve estimation robustness over time.

Joint estimation of SOC and SOH is a powerful approach to address aging effects comprehensively. Dual estimation architectures simultaneously track SOC and SOH by integrating capacity fade and resistance growth models into the estimation framework. One common method employs dual extended Kalman filters (DEKF), where one filter estimates SOC while the other updates SOH-related parameters such as capacity and resistance. The two filters operate in parallel, exchanging information to refine both estimates. Another approach uses particle filters to handle the nonlinearities and uncertainties inherent in aged battery behavior. These methods enable real-time tracking of both SOC and SOH without requiring periodic offline calibration.

Coupled estimation architectures take joint estimation further by explicitly modeling the interdependence between SOC and SOH. These methods recognize that capacity fade and resistance growth are not independent of SOC; for example, high SOC operation may accelerate certain degradation mechanisms. Coupled estimators use electrochemical or semi-empirical models to capture these interactions, improving long-term estimation accuracy. Some advanced implementations incorporate degradation physics into the state-space model, allowing the BMS to predict how aging will progress under different usage patterns.

Validation of aging-aware SOC estimation methods relies on accelerated aging test data and field performance metrics. Accelerated aging tests subject batteries to elevated temperatures, high charge/discharge rates, or deep cycling to simulate years of degradation in a compressed timeframe. These tests provide controlled datasets for tuning and verifying SOC estimation algorithms under various aging conditions. Field data from deployed systems, though slower to accumulate, offers real-world validation by capturing the combined effects of diverse operating conditions and usage patterns. Cross-validation between lab and field data ensures that SOC estimation methods remain accurate across different environments.

Performance metrics for validation include mean absolute error (MAE) and root mean square error (RMSE) between estimated and ground-truth SOC values. Ground truth is typically established through full discharge tests or reference measurements under controlled conditions. For SOH estimation, capacity fade and resistance growth predictions are compared against direct measurements taken during periodic characterization tests. Long-term tracking of these metrics across multiple aging stages confirms the robustness of the estimation approach.

In summary, aging-aware SOC estimation requires sophisticated methods to account for capacity fade, resistance growth, and evolving voltage-SOC relationships. Joint estimation of SOC and SOH through dual or coupled architectures provides a comprehensive solution, while validation using accelerated aging and field data ensures reliability. As battery systems continue to demand higher accuracy over extended lifetimes, these advanced estimation techniques will play an increasingly vital role in BMS design.