Accurately estimating the state of charge (SOC) of lithium-ion batteries has become a critical challenge in advancing battery management systems for the electric vehicle industry. Similarly, assessing the state of health (SOH) is equally crucial for diagnosing battery health and providing timely maintenance notifications in real-world applications. To address these challenges, extensive research has focused on two primary approaches: 1) developing battery models combined with filter design techniques and 2) leveraging machine learning algorithms to uncover complex patterns and correlations. In this paper, we propose an innovative approach for estimating SOH using ohmic resistance in each usage cycle. Additionally, we introduce an SOCr estimator that is trained in the first usage cycle and can be effectively applied to subsequent cycles by compensating for variations in the SOH estimate. The proposed method also integrates a health indicator derived from the estimator, enabling reliable diagnosis of the battery’s health status. This facilitates proactive management and helps maintain optimal battery performance. Experimental results demonstrate that the proposed method significantly enhances the accuracy of SOCr estimation across usages cycles. Furthermore, it enables early detection of a battery’s deterioration into an unhealthy status, allowing for timely interventions and maintenance to sustain performance and reliability.
