Prediction of indoor temperature

Prediction model logic

ClimApp allows its users to predict the indoor temperature after providing some information about the environment they are in. The calculations to predict the temperature are performed on the DTU ClimApp server. When the user chooses to enable the indoor mode and has set all the needed parameters, an API call to the server triggers the prediction engine, which takes in the parameter and returns the predicted temperature in degrees Celsius.

• API call: getIndoorPrediction
• Retrieved parameters: rho, sr, tao, wo, trv, cy, fa, no (in this order, when sent to API)

Input by users

• wo - windows open (0/1)
• trv - thermostat level (-1-6) (closed to fully open, currently in app from 1-5)
• rho - relative humidity outside (%)

Retrieved from weather service

• tao - air temperature outside (degrees celcius) optimally 3 temperature values now and two current
• sr - solar radiation (W/m^2)

Default values

Agreed upon with Jørn Toftum

• cy - construction year (1920-2010)
• fa - floor area (5-200m^2)
• no - number of occupants (1-5 people)

Images of indoor mode

• Clearly marked with text indoor
• Location and station information has been removed (could still be relevant to show)
• Swipe is not possible as we are only working with the current situation
• Images and information underneath temperature is adjusted to match indoor metrics
• All content is based on user set values