Research on Data-Driven Comprehensive Battery State-of-Health Evaluation Strategy for Roadside Energy Storage System in Electric Transportation

For roadside energy storage systems in electric transportation, assessing battery health status is crucial. This paper utilizes charging and discharging data samples from base stations and power stations under different operating conditions to present the variation of battery health status at different cycle counts. It also examines the current fluctuation and voltage changes of four batteries in the battery pack over a three-day working period. Relevant features of battery charging and discharging are extracted, and battery health status is analyzed through clustering using K-means++ and a comprehensive scoring model, resulting in four levels: excellent, good, fair, and poor. By comparing the two methods, it is found that the comprehensive scoring model performs better than Kmeans++. In summary, the research findings effectively reflect the battery’s charging and discharging capabilities, facilitating the assessment of energy storage level in charging facilities.