Published September 11, 2026 | Version v1

Airflow Distribution Datasets for ANN Training Using Adaptive Sampling and Latin Hypercube Sampling in Naturally Ventilated Spaces

  • 1. ROR icon Silesian University of Technology

Description

This dataset contains the databases used for the training and evaluation of artificial neural network (ANN) models developed for the prediction and deep interpolation of air distribution in naturally ventilated spaces. The data were generated from simulations performed for different combinations of reference wind speed and wind direction. The resulting databases contain flow-field variables recorded at spatial probe locations and were used to develop and assess the ANN and transfer learning approaches described in the associated publication.

Files

All_Probe_Results _30sim_LHS.csv

Files (194.9 MB)

Name Size
md5:1c6a601120b63875a346433aa49bcb8a
24.8 MB Preview Download
md5:cccf2ebebf409bd84ee5a90a9e036218
13.9 MB Preview Download
md5:3e979009580f1a4397f03af0077944c8
15.3 MB Preview Download
md5:9dae8763d7a32b2e84d54eb0c9af6fd6
18.1 MB Preview Download
md5:3c73dbb5fe90fe036f738ff4edf36fe7
19.9 MB Preview Download
md5:57dbe8108ef931ce5131f20c7b374fc9
22.6 MB Preview Download
md5:9fa1201e72cc73fc8d8bba8ed42b9809
38.8 MB Preview Download
md5:b79f3344725dc9ca4c2e1d665cf976a8
41.4 MB Preview Download
md5:f0bb0130cec8b1c898ca949f820d36f9
4.3 kB Preview Download

Additional details

Funding

National Science Centre
Digital human-building avatar for energy efficient building design and operation with advanced thermal occupant model 2022/47/I/ST8/02272