OpenSynth/tud-glass
收藏资源简介：
GLASS数据集包含大型客户智能电表的合成日负荷曲线，由流匹配生成模型生成。每行数据代表一个合成样本在单日内的96个功率值，以15分钟为分辨率（单位：千瓦），并受客户类别、发电类型、场景和消费/发电水平条件约束。该数据集是OpenSynth项目的一部分，旨在通过民主化合成能源数据加速全球能源系统脱碳，基于Liander N.V.的大型客户智能电表数据训练生成，覆盖荷兰能源法定义的“大消费者”（电网连接容量在60至160千瓦之间）。数据集包含43,184,437行和104列（8个元数据列和96个功率列），时间范围从2022年1月1日到2024年12月31日。它涵盖四种场景：纯消费、太阳能发电、陆上风力发电和混合净消费，涉及26个客户类别（包括消费类别和发电类别），每月分为四个水平（低、中低、中高、高），每个条件生成100-150个有效样本。重要注意事项包括风能数据仅限冬季月份（10月至2月），某些条件下生成样本多样性较低，以及数值噪声导致的符号约束处理。
This dataset contains synthetic daily load profiles for large-customer smart meters, produced by a flow-matching generative model. Each row represents a single day of a single synthetic sample: 96 power values at 15-minute resolution (in kW), conditioned on customer category, generation type, scenario, and consumption/generation level. It is part of the OpenSynth project, which democratizes synthetic energy data to accelerate the decarbonization of global energy systems, generated using smart meter data from Liander N.V. for large customers (grid connection capacity between 60 and 160 kW). The dataset has 43,184,437 rows and 104 columns (8 metadata and 96 power columns), covering dates from 2022-01-01 to 2024-12-31. It includes four scenarios: consumption_only, generation_zon (solar), generation_wopl (wind_on_land), and hybrid_net_consumption, with 26 customer categories (consumption and generation types), four levels per month (low, mid_low, mid_high, high), and 100-150 valid samples per condition. Key notes include limited wind data coverage (October-February only), low variance in some conditions, and sign clamping applied to handle numerical noise.
