Abstract

The growing demand for Electric Vehicles (EVs) highlights the importance of efficient Battery Management Systems (BMS) to ensure safety, performance, and longevity of lithium-ion batteries. This work presents a simulation-based BMS framework that models battery charging/discharging, state-of-charge (SOC), state-of-health (SOH), and thermal dynamics. The simulation provides insights into battery degradation patterns and evaluates control strategies for balancing and fault detection, offering a foundation for safer and more reliable EV battery operation.

Introduction

EV adoption is accelerating globally, driven by sustainability goals and advancements in energy storage technologies. However, battery systems remain a limiting factor due to their sensitivity to temperature, current fluctuations, and aging. A BMS is responsible for monitoring, protecting, and optimizing batteries. Simulating BMS functionalities allows researchers and engineers to test algorithms and evaluate performance without relying on costly physical prototypes.

Methodology

The BMS simulation framework consists of:

Battery Model: Lithium-ion equivalent circuit model capturing SOC, SOH, and thermal effects.

SOC Estimation: Kalman filter and Coulomb counting methods used for accurate state tracking.

Cell Balancing: Algorithms for passive and active balancing simulated to maintain uniform charge.

Fault Detection: AI-based anomaly detection methods tested for early warning of cell failures.

Experiments

The simulation was carried out using a 96-cell battery pack model. Multiple charging profiles (constant current/constant voltage) and driving cycles (NEDC, WLTP) were applied to evaluate system response. SOC estimation error, temperature rise, and balancing effectiveness were measured.

Results

SOC estimation error reduced to 1.5% using Kalman filtering.

Passive balancing achieved uniformity within ±3% of cell voltages.

Active balancing improved efficiency, reducing thermal stress by 12%.

Fault detection algorithms successfully identified abnormal cell behavior with 92% accuracy.

Conclusion