MODERATE Synthetic energy consumption profiles


FAIR Overall Score

73%

Abstract

Synthetic dataset with 500 users hourly profile of energy metering data. The original training dataset contains both residential and non residential buildings, from 5000 users. Meaning of the columns: A+: Active energy consumed from the grid (Wh) A-: Active energy exported to the grid (Wh) R1: Consumed inductive reactive energy R2: Generated inductive reactive energy R3: Consumed cpacitive reactive energy R4: Generated capacitive reactive energy NOTE: The two variants "cpu" and "xpu" do not differ statistically from each other. They are both result obtained by MOSTLY.AI library running on a local instance, run with pytorch in the CPU (multi-thread) mode, vs a XPU pytorch run. Specific statistical metrics shows that the synthetic dataset is significantly different from original profiles. While daily and weekly patterns are maintained from the synthetization process set up, in this first release the seasonal and monthly patterns are not represented in the synthetic dataset.

Citation

Landing page / DOI

Creative Commons Attribution 4.0 International

Contact for metadata

Accademia Europea


FAIR Overall Score: 73%

Principle Score Earned Level
Findable 57% 4 of 7 moderate
Accessible 100% 7 of 7 advanced
Interoperable 67% 4 of 6 moderate
Reusable 67% 4 of 6 moderate

Assessed 2026-07-22 (metrics 0.8)

Evaluated by F-UJI web service: Anusuriya Devaraju, & Robert Huber. (2020). F-UJI - An Automated FAIR Data Assessment Tool. Zenodo. https://doi.org/10.5281/zenodo.6361400


No docs sources are available.