Downloads: 1 | Views: 65 | Weekly Hits: ⮙1 | Monthly Hits: ⮙1
Research Paper | Computer Science | Nigeria | Volume 12 Issue 12, December 2023 | Rating: 4.9 / 10
A Review of Artificial Intelligence-Based Prognostic and Health Management Systems for Lithium - In Batteries in Electric Vehicles
James Oladimeji [3] | Olushola Ogunniyi [2]
Abstract: This paper presents a comprehensive review of 20 contemporary papers from the last 10 years, focusing on the use of artificial intelligence (AI) in electric vehicle (EV) battery management systems and the assessment of battery degradation status. Lithium-ion batteries are critical components of EVs, and ensuring their efficient operation and reliability is crucial for widespread adoption. Prognostic and health management (PHM) systems integrated with AI techniques have emerged as promising solutions for monitoring, diagnosing, and predicting the health status and remaining useful life of batteries. The review covers various aspects of AI-based PHM systems for lithium-ion batteries in EVs. It begins by exploring state-of-charge estimation, where studies have employed deep neural networks, recurrent neural networks, convolutional neural networks, and particle filtering techniques to enhance estimation accuracy. Additionally, the paper investigates charging and discharging algorithms, leveraging reinforcement learning and Gaussian process regression, among others, to optimize energy management in EVs and improve battery remaining useful life prediction. Furthermore, the review delves into the application of fuzzy logic-based battery management systems, dynamic ensemble models, and dual filters, demonstrating how they improve EV autonomy and enhance battery state of charge estimation. Additionally, it explores model-order reduction techniques, health-conscious kernel adaptive filtering and adaptive extended Kalman filters to analyze battery internal electrochemical transfer functions, predict remaining useful life, and estimate state-of-charge more accurately. Moreover, the review discusses data- driven approaches and Gaussian process regression for battery life prediction, as well as long short- term memory neural networks for battery state-of-health estimation. It also highlights the use of online sequential extreme learning machines for remaining useful life estimation, emphasizing their potential for real-time applications. Through this comprehensive review, the paper underscores the significant advancements made in the past decade concerning AI-based PHM systems for lithium- ion batteries in EVs. The findings highlight the potential of AI techniques in improving battery health management, optimizing charging and discharging strategies, and extending battery lifespan. Researchers, engineers, and stakeholders interested in harnessing the full potential of AI in EV battery prognostics and health management will find this research review a valuable resource.
Keywords: Artificial Intelligence (AI), Prognostic and Health Management (PHM), Lithium-ion Batteries, Electric Vehicles (EVs), Battery Management Systems
Edition: Volume 12 Issue 12, December 2023,
Pages: 345 - 355