Upgrade of the FER skyscraper smart building systems with Python application for filling the missing meteo and forecast data from the backup data


Project abstract

The crucial parts of LARES Living-lab on model predictive control covering the temperature control in all zones in FER skyscraper are meteorological station placed on the building rooftop and weather forecast support by Croatian Meteorological and Hydrological Service. The meteorological measurements are used to identify mathematical models of the building ant to train the neural network models for predicting different building-related variables such as PV system electrical energy production or overall electrical and thermal energy consumption of FER skyscraper while weather forecast is used to calculate the predictions. All relevant meteorological measurements and forecast are stored in the database which acts as a backbone of the Living-lab algorithms. In case of the communication problems with the database the relevant data are stored locally to prevent the losses of the data.

The implementation of the project is divided into the following mini projects. They could be studied independently (only certain mini project) or sequentially (mini project by mini project), based on the students’ preferences.


Mini projects

  • To have all historical data in the database it is required to develop a Python script for checking if during the selected time period passed to the script as input parameter measurement of forecast data are missing. If missing data are detected it is required to collect the missing data from backup storage and put the data in the right places in the Living-lab database. The output of the script is a detailed list of intervals with missing data, status of the data (filled from back-up storage or still missing) and communication problems frequency statistics. The testing and validation of the developed script performance will be performed on a back-up database instead of real-time database. Once the performance is verified, the script could be executed to fill the missing data in real-time on-line database of the Living-lab system. If the script is well developed it will become a part of Living-lab routines.


  • As an upgrade of the first mini project, for interested students, there is a possibility to continue working with the meteorological data by participating in planned additional upgrades of the Living-lab routines. The existence of the local building weather station enables the opportunity for improvement of the weather forecast for a specific building location. The local meteorological measurements could be used to eliminate the systematic error due to the fact that the building is not situated directly at the meteorological station site as well as due to the environment of the building itself [1].  



Additional information

  • Number of students: 1
  • Keywords: smart buildings, FER skyscraper smart control, weather forecast, meteorological station, database, smart software solutions

[1] F. Oldewurtel, A. Parisio, C. N. Jones, D. Gyalistras, M. Gwerder, V. Stauch, B. Lehmann, and M. Morari. Use of model predictive control and weather forecasts for energy efficient building climate control. Energy and Buildings, 45:15-27, 2012.