AI driven approaches in battery thermal management systems for prediction and multi objective optimization: a comprehensive literature review

Abstract Battery Thermal Management Systems (BTMS) play a critical role in ensuring the safety, performance, and lifespan of lithium-ion batteries in electric vehicles; however, conventional empirical, numerical, and physics-based approaches often struggle to handle complex thermal behaviors, nonlinear interactions, and real-time operational demands. This limitation highlights the need for intelligent, adaptable, and computationally efficient prediction and optimization frameworks. The objective of this study is to critically review and analyze recent artificial intelligence (AI)-driven approaches for BTMS, with particular emphasis on temperature prediction, thermal control, and multi-objective optimization. The methodology involves a comprehensive review of machine learning and deep learning techniques, including linear regression, support vector regression, random forest, artificial neural networks, long short-term memory networks, and convolutional neural networks, alongside optimization methods such as Response Surface Methodology, Genetic Algorithms, NSGA-II, Whale Optimization Algorithm, and other emerging techniques. The analysis reveals that AI-based models significantly enhance prediction accuracy, reduce computational cost, and enable efficient multi-objective trade-offs between cooling performance, energy consumption, safety, and design constraints. Furthermore, hybrid AI-physics models, explainable AI, digital twins, and edge-computing-enabled frameworks emerge as promising directions for real-time and scalable BTMS implementation. Overall, the reviewed studies demonstrate that AI-driven prediction and optimization can substantially improve thermal uniformity, mitigate thermal runaway risks, and support the development of safer, more efficient, and sustainable battery systems for next-generation electric vehicles.