Real time SOC estimation for Li-ion batteries in Electric vehicles using UKBF with online parameter identification

Area of Science:

• Electrical Engineering
• Materials Science
• Electrochemistry

Background:

• Lithium-ion batteries are vital for the EV industry, demanding precise State of Charge (SOC) estimation for optimal performance.
• Existing SOC estimation methods often use discrete-time dynamics, limiting accuracy.
• Accurate battery modeling is essential for reliable electric vehicle operation.

Purpose of the Study:

• To propose an accurate State of Charge (SOC) estimation method for lithium-ion batteries.
• To address the limitations of discrete-time dynamics in current SOC estimation approaches.
• To enhance the performance and precision of lithium-ion battery management systems.

Main Methods:

• Implementation of the Thevenin 2RC battery model to capture nonlinear battery dynamics.
• Integration of the Unscented Kalman Bucy Filter (UKBF) for SOC estimation using noisy voltage and current measurements.
• Simulation in Matlab Simulink for comparative analysis with Extended Kalman Filter (EKF) and Unscented Kalman Filter (UKF).

Main Results:

• The Unscented Kalman Bucy Filter (UKBF) demonstrated superior accuracy in SOC estimation compared to EKF and UKF.
• The proposed UKBF approach achieved a significantly low Root Mean Square Error (RMSE) of 0.003276.
• The UKBF effectively handles the nonlinearity of the Thevenin 2RC model and measurement noise.

Conclusions:

• The Thevenin 2RC model combined with the UKBF provides a highly accurate method for lithium-ion battery SOC estimation.
• This continuous-time dynamics approach offers a significant improvement over existing discrete-time methods.
• The UKBF is a robust filter for precise battery state estimation in electric vehicles.