Predictive Thermal Comfort Engine API

Overview

The Predictive Thermal Comfort (PTC) Engine API is intended to deliver comfort assessment calculations to the user. The API accepts various inputs as payload. Please refer to the swagger documentation and testing page here . The API response provides a link to a zip archive containing the results of the calculations as well as a log file that the user can read to see if the calculation processes performed as expected.

For a specific building case study:
• The Service reads data from a csv input file (input.csv)
• The Service takes in user inputs for air velocity, metabolic rate, and clothing insulation
• The Service reads user inputs related to the desired thermal comfort calculation method(s) to be performed, to include:
  • PMV/PPD Fanger's model
  • Adaptive thermal comfort model
  • ML machine learning model trained on the ASHRAE Global Thermal Comfort Database II (kaggle.com).
• The Service processes the data and performs thermal comfort evaluations for the desired calculation method(s)
• The Service writes results of calculations/predictions in various output files and expose the URL link to a cloud repository to download a zip archive with the main results and log file.

To get started with the PTC API Service please follow the step by step instructions for testing the API below:

• Link for Swagger documentation: https://slepc-comfort-api.saas.iesve.cloud/swagger/
• Please make sure the site is secure and there is a lock symbol in the address bar if not prefix https:// before the link
• In the "Servers" dropdown, select " https://slepc-comfort-api.saas.iesve.cloud/ " as the server
• Click on the "Try it out" button
• Select a csv input file to upload
• Fill in the values for other user input fields
• Please press the "Execute" button in order to run the API
• The response of the API provides a "Download file" URL link to a cloud repository to download a zip archive with the main results and log files.

The service requires the user to input the following:

• Input csv file containing all of the data to be evaluated, including the following variables:
• airTemp (°C): The indoor (dry bulb) air temperature
• meanRadTemp (°C): The indoor mean radiant temperature
• relativeHumidity (%): The indoor air relative humidity
• runningMeanTemperature (°C): The outdoor air running mean temperature, which can be calculated at the beginning of each day by giving a weight of "w %" to the average outdoor air dry bulb temperature for the previous day and a greater weight of "(1-w) %" to the running mean temperature for the previous day, so that adaptation to the outdoor climate is taken into account
• outdoorMonthlyAirTemp (°C): The outdoor monthly average (dry bulb) air temperature
• Example input csv file (note column/variable names are case sensitive):

  date	time	airTemp	meanRadTemp	relativeHumidity	runningMeanTemperature	outdoorMonthlyAirTemp
  06/01/2023	00:00:00	16	16	47	6	8
  06/01/2023	01:00:00	17	19	48	11	8
  06/01/2023	02:00:00	25	23	66	4	8
  06/01/2023	03:00:00	19	20	73	9	8
  06/01/2023	04:00:00	16	16	47	6	8

• air_velocity (m/s): The "Air Velocity" variable represents the velocity of the air within the space. e.g. 1 m/s
• metabolic_rate (met): The "Metabolic Rate" variable refers to the rate at which an individual's body generates heat or energy.e.g. 1 met
• clothing_insulation_level (clo): The "Clothing Insulation Level" variable describes the thermal insulation provided by the clothing worn by the user.e.g. 1 clo
• It is possible to include different calculation methods in the thermal comfort predictions. The user can choose from one to three of the following methods: PMV/PPD, Adaptive, ML. If a selection is not made the results file returned will not contain any calculations.
• Each calculation method requires specific data variables to be included within the input csv file, so the input csv file does not need to include all the data variables, depending on the calculation method(s) selected, as follows:
• PMV/PPD requires the indoor (dry bulb) air temperature (airTemp), the indoor mean radiant temperature (meanRadTemp), and the indoor relative humidity (relativeHumidity)
• Adaptive requires the indoor (dry bulb) air temperature (airTemp), the indoor mean radiant temperature (meanRadTemp), and the outdoor air running mean temperature (runningMeanTemperature)
• ML requires the indoor (dry bulb) air temperature (airTemp), the indoor relative humidity (relativeHumidity), and the outdoor monthly average (dry bulb) air temperature (outdoorMonthlyAirTemp)
• Each calculation method requires values of input data variables, and output results, to be within specific standard applicability limits (if values are outside standard applicability limits, the calculation method will return an empty result value in the output files), as follows:
• PMV/PPD requires 10 °C < airTemp < 30 °C, 10 °C < meanRadTemp < 40 °C, 0.8 clo < clothing_insulation_level < 4.0 clo, and -2 < PMV < 2
• Adaptive requires 10 °C < airTemp < 40 °C, 10 °C < meanRadTemp < 40 °C, 10 °C < runningMeanTemperature < 33.5 °C, and 0 m/s < air_velocity < 2 m/s

Try it out

To use the Predictive thermal comfort service, please click here