A Framework for Automatic Interpretation of Electric Vehicle Industry Policies Based on LLMs

The electricity demand of power-intensive industries is influenced by national policies. Systematically assessing the potential impact of such policies not only helps industries adapt to policy changes but also provides scientific reference for policymakers. However, traditional policy analysis methods often struggle with the complexity of policy documents, resulting in low efficiency and insufficient accuracy. This study proposes a policy interpretation framework based on Large Language Models that automatically analyzes policy documents, constructs a multilevel policy evaluation model, and quantitatively assesses the impact of policies on the structure of electricity demand within industries. Using this analytical framework, this study examines 39 electric vehicle industry policy documents released by Zhejiang Province, China over the past five years to generate a multilevel rule model. The results demonstrate that the framework effectively captures key factors influencing electricity demand in the electric vehicle industry and provides quantified outcomes, showing significant advantages in improving the efficiency of policy analysis.