Application of Dynamic Cell Resistance for determination of state of charge

Li-ion batteries have found an ever-increasing role in automotive, space and marine industries, and safety aspects became a top priority, particularly for large size batteries. Therefore, there has been an increasing need for accurate and reliable monitoring of the battery in real-time. The battery is a dynamic system and its parameters are changing with time. They are also very dependent on the operation history of the battery. Hence, the impact of aging needs to be effectively addressed within any monitoring scheme. This work tries to bring a new dimension to battery monitoring by introducing the Dynamic Cell Resistance where it is closely related to battery cycling history and the state of charge of the battery. This parameter is modeled versus state of charge using a Group Method of Data Handling (GMDH) neural network.