Real-Time Remaining useful Life Estimation and Dynamic Pricing for Electric Vehicle Battery Swapping Stations using Machine Learning

C Raja Sekhar1, Arigela Ganesh2,Kolleti varshini3,Balasa Chenchu Aalaya4,Katari Tharun Teja5
1Assistant Professor, Dept of Information Technology, SV College of Engineering, Tirupati, India.
2B.Tech, Dept of Information Technology, SV College of Engineering, Tirupati, India.
3B.Tech, Dept of Information Technology, SV College of Engineering, Tirupati, India.
4B.Tech, Dept of Information Technology, SV College of Engineering, Tirupati, India.
5B.Tech, Dept of Information Technology, SV College of Engineering, Tirupati, India.
Email: 1 rajasekhar.ch@svce.edu.in, 2 ganesharigela147@gmail.com, 3kolletivarshini@gmail.com,
4balasaaalaya2004@gmail.com, 5tharuntejakatari05@gmail.com
Corresponding Author* : C Raja Sekhar

Abstract-An electric vehicle (EV) battery is a rechargeable battery that powers the electric motors of battery electric hybrid electric vehicles (BEVs)  or vehicles (HEVs).These batteries are typically high-capacity lithium-ion types,optimized for high power-to-weight ratio andenergydensity to provide sufficient driving range and performance Electric vehicle (EV) battery swapping systems currently face challengesof unfair pricing because fixed or subscription models charge users the same regardless of battery health, ignoring differences in battery degradationand performance. These systems lack real-time  estimation of the remaining useful life (RUL) and do notdynamically assess critical battery parameters such as discharge time and voltage, resulting in inefficiencies and poor user experience. A proposed solution integrates an XGBoostmachine learning algorithm for accurate, real-time RUL prediction with a dynamic pricing strategy that adjusts prices based on detailed battery health metrics. This approach ensures fairer pricing aligned with battery quality, enhances user satisfaction by enabling equitable battery swaps,and improves operational efficiency through optimized battery allocation. The system is also computationally efficient and has low memory requirements for practical deployment, with potential future enhancements using hybrid models,advanced feature engineering, and large-scale validation across diverse battery types and environments Keywords: Battery electric vehicles, batterydegradation, hybrid models, real-time RUL prediction