Merging Large Language Models and Battery Physics for User-Aware Electric Vehicle Driving Management
In the authors' words
Electric vehicle (EV) battery performance is strongly coupled with driver behavior, yet human intent is typically expressed semantically rather than numerically. This paper proposes a hybrid physics-artificial intelligence framework that integrates a Large Language Model (LLM) as a high-level behavioral reasoning layer within a physics-driven supervisory architecture. The LLM interprets textual user intent and structured battery feedback to generate bounded behavioral parameters that shape a discharge current envelope. A physics-driven safety filter then enforces physical safety constraints before computing feasible velocity recommendations. Lyapunov-based analysis establishes bounded recommendation error under battery model and prompt inaccuracies. Simulation results demonstrate adaptive, user-aware operation without compromising physical safety. The proposed reasoning-enforcement architecture provides a principled pathway for safe AI integration in EV energy management.
Appeared: Thursday, September 24. arXiv. Preprint, not yet peer-reviewed.