Computational Method for Optimal Electrolyte Screening Using Bayesian Optimization and Physics Based Battery Model

Lithium-ion batteries (LIBs) powering electric vehicles and large-scale energy storage depend significantly on the composition of liquid electrolyte for optimal performance. We propose a framework coupling Bayesian optimization and physics based battery models to identify electrolytes optimal for specific set of requirements such as less capacity fade and internal heating etc. Our approach is validated through multiple case studies, demonstrating the framework’s efficacy in optimizing electrolyte properties. Additionally, we introduce a deviation index to quantify the proximity of the optimal electrolyte to those in a predefined database. With adaptability to various LIB metrics and battery chemistries, it provides a systematic and efficient approach for screening electrolytes based on system-level performance using physics-based models, contributing to advancements in battery technology for sustainable energy storage systems.