用视觉大模型分析建筑能耗数据，发现异常并给出节能建议
Extracting Actionable Insights from Building Energy Data using Vision LLMs on Wavelet and 3D Recurrence Representations
- 将时序数据转为3D图像，让大模型像看图一样分析能耗
- 在真实数据上实现0.0952的低损失，优于直接处理原始数据
- 适合能源管理、智能建筑领域研究人员和工程师
由于建筑能耗数据具有非线性和多尺度特性，从中提取可操作的洞察仍具挑战。为此，我们提出一种框架，通过在3D图形表示上微调视觉语言大模型（VLLMs）来解决该问题。该方法利用连续小波变换（CWTs）和递归图（RPs）将一维时序数据转换为3D表示，捕捉时间动态并定位频率异常。这些3D编码使VLLMs能够视觉化解读能耗模式，检测异常，并生成节能优化建议。我们在真实建筑能耗数据集上验证了该框架，微调后的Idefics-7B VLLM在阿联酋沙迦大学数据集上，使用CWTs时验证损失为0.0952，使用RPs时为0.1064，优于直接在原始时序数据上微调的0.1176，显著提升异常检测性能。本工作连接了时序分析与可视化，提供了一种可扩展且可解释的能源分析框架。
原文摘要 · Abstract (English)
The analysis of complex building time-series for actionable insights and recommendations remains challenging due to the nonlinear and multi-scale characteristics of energy data. To address this, we propose a framework that fine-tunes visual language large models (VLLMs) on 3D graphical representations of the data. The approach converts 1D time-series into 3D representations using continuous wavelet transforms (CWTs) and recurrence plots (RPs), which capture temporal dynamics and localize frequency anomalies. These 3D encodings enable VLLMs to visually interpret energy-consumption patterns, detect anomalies, and provide recommendations for energy efficiency. We demonstrate the framework on real-world building-energy datasets, where fine-tuned VLLMs successfully monitor building states, identify recurring anomalies, and generate optimization recommendations. Quantitatively, the Idefics-7B VLLM achieves validation losses of 0.0952 with CWTs and 0.1064 with RPs on the University of Sharjah energy dataset, outperforming direct fine-tuning on raw time-series data (0.1176) for anomaly detection. This work bridges time-series analysis and visualization, providing a scalable and interpretable framework for energy analytics.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。