Work Description

Title: Data Files for Real-Time Energy Disaggregation of a Distribution Feeder's Demand Open Access Deposited

http://creativecommons.org/licenses/by/4.0/
Attribute Value
Methodology
  • This data set corresponds to that used within "Real-Time Energy Disaggregation of a Distribution Feeder's Demand Using Online Learning", which is available on arXiv at ( https://arxiv.org/abs/1701.04389). The contents were derived from original data from the Pecan Street Dataport, the Pacific Gas & Electric Company, and NOAA. Please refer to the paper for details on the data processing and relevant citations for the original data sources.
Description
  • This data set contains the relevant time series for constructing and testing electricity load models within the related paper. The files within are a '.mat' file that contains the data and a 'readme.txt' file detailing the contents of the data.
Creator
Depositor
  • gsledv@umich.edu
Contact information
Discipline
Funding agency
  • National Science Foundation (NSF)
ORSP grant number
  • #ECCS-1508943
Keyword
Date coverage
  • 2015-05-02 to 2015-08-18
Citations to related material
Resource type
Last modified
  • 04/02/2020
Published
  • 07/24/2017
Language
DOI
  • https://doi.org/10.7302/Z2668BBC
License
To Cite this Work:
Mathieu, J., Balzano, L., Ledva, G. (2017). Data Files for Real-Time Energy Disaggregation of a Distribution Feeder's Demand [Data set]. University of Michigan - Deep Blue. https://doi.org/10.7302/Z2668BBC

Relationships

Files (Count: 2; Size: 49.3 MB)

The data contained in this directory corresponds to the data used within the publication
"Real-Time Energy Disaggregation of a Distribution Feeder’s Demand Using Online Learning",
which is available on arXiv at https://arxiv.org/abs/1701.04389. Please refer to this paper
for any relevant details on the processing of the data from its original sources, which are
also cited wtihin the paper.

The data file dataForFeederLevelEnergyDisaggregation.mat contains two structures (explained
in detail below):

1) testingDaysData
2) trainingDaysData.

===================================

1) testingDaysData contains fields:
a) acDemand
- the aggregate AC demand signal (kW)
b) olDemand
- the aggregate OL demand signal (kW)
c) olDemandCommercialComponent:
- the commercial demand within olDemand (kW)
d) olDemandResidentialComponent:
- the residential demand within olDemand (kW)
e) outdoorTemperatureForResidentialDemand:
- the outdoor temperature (degC) that corresponds to the residential demand
f) outdoorTemperatureForCommercialDemand:
- the outdoor temperature (degF) that corresponds to the commercial demand
g) dateAndTimes:
- the date and time for each datapoint in the above fields

Each row within a-f corresponds to data for a single testing day.
Each column within a-f corresponds to a time-step within the day, where the time-step used is 1 minute.

The dateAndTimes field is a cell array, where each cell contains the date and time of each time-step for a given testing day.
Note that testingDaysData.dateAndTimes{1} corresponds to the first row of testing day data within a-f, and so on.

===================================

2) trainingDaysData contains fields:
a) acDemand
- the aggregate AC demand signal (kW)
b) olDemandResidentialComponent
- the residential demand component of the OL demand (kW)
c) olDemandCommercialComponent
- the commercial demand component of the OL demand (kW)
d) outdoorTemperatureForResidentialDemand
- the outdoor temperature (degC) that corresponds to the residential demand
e) outdoorTemperatureForCommercialDemand
- the outdoor temperature (degF) that corresponds to the commercial demand
f) dateAndTimes
- the date and time for each datapoint in the above fields
g) houseAcData
- a structure containing the fields:
i) houseIdsOnFeeder
- contains the list of house IDs included within the feeder
ii) numberOfOccurancesPerId
- the first element of numberOfOccurancesPerId corresponds to the number of times
the first house ID was included within the aggregate data, and so on.
iii) householdAcTimeSeries
- time series of the AC demand for each house ID given in houseIdsOnFeeder
- Each row is a datapoint for a one minute time-step, and each column contains
the time series for a specific house ID.
- the first element of houseIdsOnFeeder corresponds to the ID for the first
column of householdAcTimeSeries, and so on.
iv) dateStrings
- the date and time for each time-step in the householdAcTimeSeries

Each row in a-e corresponds to a datapoint for a one minute time-step, and f contains the date and time for that time-step.
Note that any NaN values were excluded from model construction.

Download All Files (To download individual files, select them in the “Files” panel above)

Best for data sets < 3 GB. Downloads all files plus metadata into a zip file.



Best for data sets > 3 GB. Globus is the platform Deep Blue Data uses to make large data sets available.   More about Globus