Loading...

Fault Diagnosis of Lithium-ion Battery Pack of Electric Vehicles Using Machine Learning Algorithms

Noori, Fatemeh | 2024

5 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 56899 (46)
  4. University: Sharif University of Technology
  5. Department: Energy Engineering
  6. Advisor(s): Moeini Aghtaei, Moein
  7. Abstract:
  8. In the current era, the urgent need to reduce the use of fossil fuels and limit carbon emissions has highlighted the importance of shifting from traditional gasoline-powered vehicles to electric vehicles. The evolving landscape and challenges in the field of electric vehicles highlight the importance of ensuring the safety and reliability of energy storage systems. Lithium-ion batteries play a central role in electric vehicle technology, requiring thorough research efforts. Notably, the detection of internal short circuit, a common issue in lithium-ion batteries, has become a focus of researchers in recent years. This study aims to investigate the internal short circuit fault in lithium-ion batteries, recognizing its pivotal role in causing catastrophic failures. Furthermore, the comprehensive consideration of battery aging and cell inhomogeneity, incorporating alterations in battery model parameters in simulations, presents an approach that aligns closely with real-world conditions. The development of an equivalent circuit model in MATLAB software, encompassing both battery cells affected by internal short-circuit fault and healthy battery cells, alongside a diverse range of cell capacities, enables a precise exploration of internal short-circuit faults across varied operational scenarios. The resulting data from the simulator model serves as input for machine learning algorithms under diverse operational scenarios. The classification outcomes of the proposed model exhibit a notable accuracy rate of 95% for the random forest method. Furthermore, a comprehensive analysis, employing various evaluation metrics, has been undertaken to compare the performance of the random forest model with two other machine learning techniques. The findings confirm the effectiveness of the random forest algorithm in detecting battery faults, indicating its capability to elevate the safety and dependability of energy storage systems
  9. Keywords:
  10. Aging Effects ; Fault Detection ; Lithium Ion Batteries ; Machine Learning ; Internal Short Circuit ; Electric Motor Vehicles

 Digital Object List

 Bookmark