DYNASTEE Datasets made available for training purposes.
The DYNASTEE network has gathered over a long period a number of well documented high-quality datasets. Several have been used regularly in the training events, like workshops, Summer Schools, etc. Producing high-quality data is a very expensive process and many man-hours are spent to carry out the experimental work and to create these data series. In most cases they have been used for training people the specific issues of dynamic analysis in building energy performance research projects but also to develop dedicated standards (CEN and ISO).
Participants to the DYNASTEE workshops and training events in the past have been using several of the available datasets and sometimes presented results in scientific articles in Journals. The DYNASTEE network has started in April 2023, in order to keep track of where the datasets are used, a registration procedure. The datasets are free to be used for educational purposes and may not be sold in any format. When used for scientific articles a proper reference is requested that should include the DYNASTEE network (www.dynastee.info ), the reference to the applied dataset and the date of download. Request for data can be addressed to contact firstname.lastname@example.org
At present three zipped data-series are available (see the list below also).
- The first data-series are simulated data and are grouped in BenchmarkTestDynMethods20.zip. This zipped folder includes the following files: Dataseries in text format and pdf (document describing two cases for self-training) and defstatest.pdf (document that contains definitions and statistical tests).
- The second data-series are real experimental data from a Round Robin test box. Minutely data of 45 sensors are made available for about a month divided into three separate periods, each with its particular heating regime. These data make it possible to perform different analysis of walls, window and space. The data-series are placed in the zipped folder zip and contains three data-series (Almeria_series16, 17 and 18) including the documentation SS20_RRbox_Instruction_document.pdf (describing the experimental setup). A recent extensive paper: dynamicanalysisapplied2epb.pdf presents the data analysis process applied to these high quality data-series.
- The third data-series are real experimental data from a well-controlled In-Situ experiment of a homogenous concrete wall. Two periods of consecutive data are available; the first one with shading of the external wall and hence the mounted sensors. In the second period the shading device has been removed. These data-series are placed in the zipped folder zip and contains two data-series (GFwallShaded.csv and GFwallNotshaded.csv), the description of the experimental setup, In_Situ_GFwall20s.pdf and a proposal for an exercise In_Situ_ExerciseWall_20s.pdf.
In addition, two extensive documents dealing with data analysis, have been made available from the IEA-EBC Annex 58 project (2011 – 2016):
- The document Guidelines_Analysis_BuildingPhysics_A58.pdf focuses on criteria that must be considered to avoid mistakes in pre-processing data and constructing candidate models.
- The document Guidelines_Analysis_StatisticalModelling_A58.pdf presents criteria for selecting the optimal method and model to analyse the available data.
DYNASTEE is testing other datasets from the list below. When they are tested and accepted, information will be placed on our website.
- Simulated data homogenous wall – available
- In-Situ data from homogenous wall – available
- In-Situ data from composition wall soon available
- Data from Round Robin box (research) – available
- In-Situ data from an air gap envelop soon available
- Data from a whole building under review
- Data from a co-heating site experiment under review
- Smart metering data work in progress
Note that DYNASTEE has started organising the next Summer School in Almeria, Spain in September 2024. Feel free to contact email@example.com to be placed on the Summer School mailing-list and you will receive further information as soon as available.