State of Charge Estimation and State of Health Diagnostic Method Using Multilayer Neural Networks

Lithium batteries are the most common energy storage devices in fields such as electric vehicles, portable devices, and energy storage systems. Continuously using the battery in a degradation state creates a fire or explosion risk. To prevent such accidents, research on a battery management system (BMS) that diagnoses the state of a battery was conducted. This study proposes a method that uses multilayer neural networks (MNN) for state of charge (SOC) estimation and state of health (SOH) diagnosis. The proposed method uses four MNN models as SOH diagnostic models and three SOC estimation models. Each SOC estimation model comprises a normal model, a caution model, and a fault model according to the learned data based on the output result of the SOH diagnostic model. From the experiments, the proposed method estimates and diagnoses SOC and SOH well.