سومین کنفرانس بین المللی مهندسی و علوم کامپیوتر و اولین کنفرانس ملی سامانه های هوشمند، کلان داده و رایانش پیشرفته

رفتن به صفحه اصلی

ارائه مجازی مقاله

کد مقاله CCES-1084

عنوان Machine Learning-Based Prediction of Lithium-Ion Battery State of Health and Optimization of Operational Conditions for Enhanced Lifespan

نویسندگان Muhammad Saeidi - Reza Ebrahimi-Kahrizsangi - Masoud Kasiri-Asgarani

چکیده Accurate prediction of lithium-ion battery State of Health (SOH) is essential for enhancing the longevity, safety, and reliability of energy storage systems. This study employs a Random Forest regression model to estimate SOH using a publicly available dataset that includes cycle-level operational parameters such as temperature, current, voltage, and cycle number. Initial analysis revealed strong correlations between SOH, remaining useful life (RUL), and battery capacity (BCt). Feature importance and partial dependence analyses identified discharge temperature, current, and voltage as the most influential factors affecting SOH, with elevated discharge temperatures having the most detrimental effect. A simulation of optimized operating conditions—featuring moderate voltages and reduced thermal and electrical loads—predicted the potential to maintain SOH at approximately 99.3%. The model achieved a high predictive accuracy with an R² of 0.9996 and mean squared error of 0.099, demonstrating its potential integration into intelligent Battery Management Systems (BMS) for real-time diagnostics and degradation-aware operational control.

فایل ها

فایل ارایه مقالات پوستری

محتوای ارائه مجازی مقاله

نظرات