Optical-Sensor-Based State-of-Charge Estimation for Lithium Iron Phosphate Batteries

This work presents a new optical-sensor-based method for state of charge (SOC) estimation in lithium iron phosphate (LFP) batteries used in electric vehicles (EVs). A fiber Bragg grating sensor is embedded inside a lab-made LFP pouch cell to provide real-time strain measurements on the anode. The optical-sensor-integrated cell is tested under the federal urban driving schedule, which emulates the dynamic duty cycle of real-world EV operations. An ensemble Gaussian-process-regression-based SOC estimation algorithm is developed by combining optical sensor measurements with conventional electrical data. The proposed method achieves a root mean square error of 1.51%, which is 2.7 times lower than that of the conventional Extended Kalman Filter-based approach. These results highlight the potential of optical-sensor-based, data-driven SOC estimation methods for LFP batteries under real-world driving cycles.