Incorporating state of charge estimation methods towards more accurate monitoring of second-life lithium-ion batteries

Although several methods are developed for estimating the state of charge of lithium-ion batteries, there is still a challenge regarding monitoring the second life of these batteries due to the expected difference in the behavior and operating conditions in their second life. Considering that each method has its advantages and disadvantages according to the application and operating conditions, and monitoring the batteries in their second life is of special importance because it is required to integrate with the battery management system to balance cells, diagnose faults and prevent overheating. Therefore, the estimation method developed in this study, by incorporating artificial neural network method and Kalman filter method with fine-tuning of the filtering process using Coulomb counting, provides a solution for more accurate online monitoring of these batteries. Aiming to get the best possible performance, considering the specificity of the second life of the batteries in terms of operating voltage, low values of expected capacity, discharge ratio and other operational parameters.