Dataset contents - Controlled artificial noise sources - Measurements - Microphones - Sound level meters - Layouts and tests - Real situations - Cano Street - Campaigns - Sound level meter measurements - Yolov8 detections   - Sargento Llagas Street - Campaign - Measurements - Solver for Windows Description of contents Controlled artificial noise sources contains the sound intensity level records and a schematic of the layouts that were tested in the model validation experiment with controlled artificial noise sources. Microphones contains the collection of audio recordings that was conducted using microphones as the receiving devices, with each tested layout being stored in a separate folder. The audio files (WAV files) and data files (CSV files) are provided for each test. Furthermore, the reference spectrum employed in the model as the spectrum emitted by the sources is provided. Sound level meters contains the records made using sound level meters as receiving devices; only the data files (in Excel format) are provided. In the case of the sound level meters, continuous recording was carried out throughout the experiment, while with the microphones, only the duration of each test was recorded. Real situations comprises all the records made during the measurement campaigns carried out in Cano Street and Sargento Llagas Street. For Cano Street, three measurement campaigns were conducted, with data files containing sound intensity level records and text files detailing the results of people detection using computer vision provided for each. Similarly, Sargento Llagas Street has one unique set of data, comprising sound intensity level records and results of people detection. However, Sargento Llagas Street data (sound intensity level + number of people) is consolidated into a single file for each day. Solver for Windows contains a Windows executable that allows the developed model to be used to predict noise levels at any position within a Street Canyon, depending on its geometry and the number of sources present. An input data file for the model and the output data files returned by the model are included as examples. How to use the model Input The first step is to create a text file containing: - geometry data of the street to be analysed, - number and location of the areas of the street where people tend to congregate, - number of people (or noise sources) in each congregation, - and the coordinates at which the noise level prediction is to be made. A text file containing the necessary input data is included in the directory as an example. The structure of each row of the input file should be as follows: 7.50 --> width of the street, in metres. 2 --> number of areas where crowds of people are likely to occur. 1 3.50 9.00 2.00 31.00 --> coordinates and size of each area where crowds form. In a separate row for each area: 2 2.00 4.00 0.00 18.50 [area_number   x-coord.   y-coord.   width   height]. The coordinates (x, y) refer to the lower left corner of each area. The x-axis is in the transverse direction of the street, the y-axis in the longitudinal direction. All data in metres. Idem for the second area. One row for each area to be defined. 1.0 1.5 1.0 --> 3 parameters, in the following order: distance at which the reference spectrum of the sources was recorded, height of the sources with respect to the ground and a parameter defining whether the third reflective surface (ground) is present or not: 1 -> exists;   0 -> does not. 3 --> number of cases raised, with different crowd densities for each area. 1 28 20 15 --> area number followed by the number of sources present in that area in each case. A value for the number of sources for each case considered: [area_number sources_case_1 sources_case_2 sources_case_3 ...] 2 7 20 12 --> idem for the second area. One row for each defined area. 30 --> number of calculation iterations with different positioning of sources within each defined area and different offset between the spectrum emitted by each source. 2 --> number of receivers or positions where the sound intensity level is to be predicted. 1 7.20 55.95 14.90 --> coordinates of the positions where the sound intensity level is to be predicted. In a separate row for each position: [receiver_number   x-coord.   y-coord.   height]. All data in metres. 2 7.20 47.95 5.45 --> idem for the second position. One row for each position to be defined. Run the model The model is executed directly in the operating system, through the system PowerShell, by sending a prompt indicating the path to the executable, the path to the input file, and the path to which the output files are to be stored. The command must conform to the following format: "path/to/model  <  path/to/input_file  >  path/to/output_file1  path/to/output_file2  path/to/output_file3" i.e.:   "Solver_20Rbis.exe  <  input_data.d  >  output.s6  output.s7  output.s8" Output The model returns three text files containing the results. The first file contains the input data, the coordinates of each source in each calculation iteration, and the prediction of the sound intensity level in each of the defined positions, for each case and for each calculation iteration. The second text file contains the solution of the half-space model (considering only one reflecting surface, the ground) at each of the defined positions, for each case considered and for each calculation iteration. The third text file contains the differences in decibels (dB) between the two aforementioned solutions.