Machine Learning and Time-Series Foundation Models for Day-Ahead Electricity Forecasting in a University Building

📄 arXiv: 2610.12638v1 📥 PDF

作者: Amey Vasulkar, Samuel Kernan Freire, Arie-Willem de Leeuw, Tadeo-Baldiri Salcedo-Rahola

分类: cs.LG, cs.AI

发布日期: 2026-10-08

💡 一句话要点

提出基于机器学习的模型以优化大学建筑的电力预测

🎯 匹配领域: 支柱九：具身大模型 (Embodied Foundation Models)

关键词: 电力预测 机器学习 时间序列 建筑管理 模型评估 气象变量

📋 核心要点

  1. 现有电力预测方法在准确性与计算需求之间存在权衡，难以满足实际应用需求。
  2. 本文提出了基于TimesFM和XGBoost的电力预测模型，结合气象变量以提高预测准确性。
  3. 实验结果显示，TimesFM模型在假期预测中表现优异，整体误差显著降低，成为强有力的基线模型。

📝 摘要（中文）

日常电力预测可以帮助建筑运营者根据预期需求安排供暖、通风等服务。选择预测模型需要在准确性、输入时间、局部训练和计算需求之间取得平衡。本文对统计、机器学习、递归神经网络和预训练的TimesFM模型进行了基准测试，使用时间验证法评估57个测试日的准确性。结果表明，TimesFM 3.0模型在加入气象变量后，整体误差最低，相较于每周季节性简单预测，平均绝对误差降低约43%。尽管XGBoost表现竞争力，但在日常比较中未能明显超越TimesFM。TimesFM在仅使用过去电力消耗时也表现良好，加入气象变量改善了假期预测，但对工作日和周末预测略有影响。TimesFM无需建筑特定的权重训练，而XGBoost则提供快速的局部拟合和推理。

🔬 方法详解

问题定义：本文旨在解决大学建筑的日常电力预测问题，现有方法在准确性和计算效率上存在不足，难以满足实际需求。

核心思路：提出结合机器学习和时间序列模型的TimesFM，利用历史电力消耗和气象变量进行预测，以提高准确性和适应性。

技术框架：整体架构包括数据收集、模型训练与验证、预测输出等阶段，采用时间验证法评估模型在不同日期的表现。

关键创新：TimesFM模型的预训练特性使其无需建筑特定的权重训练，显著降低了模型的复杂性和计算需求。

关键设计：模型设计中，选择合适的损失函数和参数设置，确保在不同工作日、周末和假期的预测准确性，同时考虑气象变量的影响。

🖼️ 关键图片

fig_0
fig_1
fig_2

📊 实验亮点

实验结果显示，TimesFM 3.0模型在加入气象变量后，整体平均绝对误差降低约43%，相较于每周季节性简单预测，显示出显著的性能提升。尽管XGBoost在局部拟合中表现竞争力，但未能在整体准确性上超越TimesFM。

🎯 应用场景

该研究的潜在应用领域包括智能建筑管理、能源调度和可再生能源集成等。通过准确的电力预测，建筑运营者可以更有效地管理能源使用，降低成本并提高舒适度。未来，该模型可扩展至其他类型建筑和更广泛的能源管理系统。

📄 摘要（原文）

Day-ahead electricity forecasts can help building operators schedule heating, ventilation, and other services around anticipated demand. Choosing a forecasting model also requires balancing accuracy with its forecast-time inputs, local training, and computational requirements. We benchmark statistical, machine-learning, recurrent neural-network, and pretrained TimesFM models for a university building. We use chronological validation and evaluate all configurations on the same 57 test days, comparing accuracy across workdays, weekends, and holidays alongside local fitting and forecasting time. Weather covariates use retrospective observations. TimesFM 3.0 with covariates achieves the lowest overall errors, reducing mean absolute error by approximately 43\% relative to weekly seasonal-naive forecasting. XGBoost remains competitive, although paired daily comparisons do not establish a clear mean-absolute-error advantage for TimesFM. TimesFM also performs strongly using only past electricity consumption. Adding covariates improves its holiday forecasts but slightly worsens workday and weekend forecasts, leaving overall mean absolute error almost unchanged. TimesFM requires no building-specific weight training, while XGBoost offers rapid local fitting and inference. The evaluated LSTM takes longer to fit without outperforming either approach. An exploratory historical evaluation reveals seasonal variation, while a separate weather-source comparison produces only small changes in the selected TimesFM configuration's accuracy. These results support pretrained TimesFM as a strong initial forecasting baseline for this building, with XGBoost as a competitive locally fitted alternative. Covariates are particularly useful when improved holiday forecasts justify the additional input preparation.