Accurate SoC estimation is critical for battery safety and cycle life. This article compares coulomb counting and extended Kalman filter approaches, explaining when each is appropriate for UAV, robotics and energy storage systems.
Coulomb Counting
The simplest method integrates current over time. It is easy to implement but drifts due to measurement noise and initial error.
Kalman Filter
The extended Kalman filter fuses current, voltage and temperature measurements with an electrochemical model, providing better accuracy under dynamic loads.
Recommendation
For high-value industrial platforms, we recommend a model-based Kalman approach combined with periodic recalibration at full charge.