Dataset name: Berkeley Energy Disaggregation Data Set (BERDS)
Data characteristics | Multivariate time series | Type of building(s) | Commercial | Date | 2012-2013 |
Type of attributes | Real and integer numbers , string | Number of buildings | 1 | Duration | One year |
Level of detail | Whole building, Appliance level | Location of building(s) | Cory Hall on the UC Berkeley campus, US | Sampling frequency | Every 20 seconds |
Access to the dataset
The data are freely available from :http://www.eecs.berkeley.edu/~maasoumy/
also more similar data from different places in the campus are available from : http://berkeley.openbms.org/map/"Data set description
The dataset include energy consumption and indoor outdoor temperature. The data are presented in four consumption classes of 1) Lighting power 2) HVAC power 3) Receptacle power, and 4) Other. The data are targeted for disaggregation purposes especially in commercial buildings. Meta data of the dataset is described in the paper mentioned here which is downloadable from the link above.
Dataset structure
The Dataset files are in csv file format. For each class of consumption three measurements data file are given which are Apparent power, true power and reactive power.Attributes
Apparent power (kVA)
True power
Reactive power (kVAR)
Paper and citation
M. Maasoumy, B. M. Sanandaji, K. Poolla, and A. S. Vincentelli, "BERDS-BERkeley EneRgy Disaggregation Data Set."