SOC estimation of lead-carbon battery based on GA-MIUKF algorithm

Area of Science:

• Battery Technology
• Electrochemical Engineering
• Computational Intelligence

Background:

• Accurate State of Charge (SOC) estimation is crucial for the performance and safety of lead-carbon batteries.
• Existing estimation methods may lack precision, particularly under dynamic operating conditions.
• Optimization of battery model parameters and filter noise is essential for enhancing SOC estimation accuracy.

Purpose of the Study:

• To propose and validate a novel SOC estimation method for lead-carbon batteries using the GA-MIUKF algorithm.
• To enhance the precision of SOC estimation by optimizing battery model and noise parameters.
• To compare the performance of the proposed GA-MIUKF method against conventional UKF and MI-UKF algorithms.

Main Methods:

• Development of an equivalent circuit model for lead-carbon batteries.
• Application of the Genetic Algorithm (GA) for global optimization of battery model parameters and MI-UKF noise variance parameters.
• Implementation and comparative analysis of the Multi-innovation Unscented Kalman Filter (MI-UKF) and Unscented Kalman Filter (UKF) algorithms.

Main Results:

• The GA-MIUKF algorithm demonstrated superior performance in SOC estimation for lead-carbon batteries.
• The proposed method achieved an average estimation error of 2.0%, outperforming standard UKF and MI-UKF algorithms.
• Optimization of parameters using GA significantly contributed to the enhanced accuracy of the MI-UKF algorithm.

Conclusions:

• The GA-MIUKF algorithm provides a highly accurate and effective approach for lead-carbon battery SOC estimation.
• The integration of GA for parameter optimization enhances the robustness and precision of the MI-UKF filter.
• This research offers a valuable contribution to improving battery management systems for lead-carbon battery applications.