Improved Accuracy of Online State of Charge Estimation for Wireless Sensor Network Nodes

Energy-aware optimizations are implemented to extend wireless sensor networks (WSNs) lifetime, which is primarily restrained by the limited capacity of the nodes' battery. Furthermore, existing WSN node's state of charge (SOC) estimation approaches lack online recalibration capability which may result in erroneous SOC values and degrades energy-aware optimizations' performance. This contributes to battery damage and shortens the sensor network's lifespan. As a solution, an online SOC estimation method with a parameter-update capability called Battery and Segmented-OCV Models' Parameter Recalibration Scheme with Kalman Filter (BSPR-KF) is developed and implemented. Simulations show that BSPR-KF provides more accurate SOC estimates in the presence of measurement noise compared to online SOC estimation methods with predetermined parameters namely: 18th-order Polynomial OCV-CCM (POC), 18th-order Polynomial with Extended Kalman Filter (POC-EKF) and Combined Model with Extended Kalman Filter (CM-EKF).