An accurate battery state of health estimation method easy to imlement after charging

Abstract

While lithium-ion batteries are widely deployed to large-scale applications, such as electric vehicles and stationary energy storage plants, the gradual degradation of batteries impose significant influence on their safety and efficiency during operation. Thus, an accurate method needs to be developed to estimate the battery state of health (SOH). Existing SOH estimation methods in recent literature mainly fall into two categories: model-based and data-driven methods. Model-based methods attempt to expand the original battery model by introducing components reflecting the influence of various factors affecting the battery degradation. On the other hand, data-driven methods aim to characterize the relation between battery SOH and various health indicators by machine learning. Once the training process has been completed, such data-driven methods are less complex to implement than model-based methods, making them more promising for practical applications.