📄 中文摘要
该研究提出了一种基于大型语言模型（LLMs）的少量样本分类框架，用于预测次日实时电价是否会出现剧烈波动。该方法将系统状态信息进行汇总，包括电力需求、可再生能源发电、天气预报和近期电价，形成一组统计特征，并将其格式化为自然语言提示，输入给LLM，同时附上通用指令。模型随后评估次日成为波动日的可能性，并报告置信度分数。通过使用德克萨斯州电力市场的历史数据，展示了该少量样本方法在性能上可与监督机器学习模型相媲美。
基于少量样本的大型语言模型框架用于电力市场的极端日分类
📄 English Summary
A Few-Shot LLM Framework for Extreme Day Classification in Electricity Markets
This research proposes a few-shot classification framework based on Large Language Models (LLMs) to predict whether the next day will experience spikes in real-time electricity prices. The approach aggregates system state information, including electricity demand, renewable generation, weather forecasts, and recent electricity prices, into a set of statistical features formatted as natural-language prompts and fed to an LLM along with general instructions. The model then assesses the likelihood of the next day being a spike day and reports a confidence score. Using historical data from the Texas electricity market, the few-shot approach demonstrates performance comparable to supervised machine learning models.