Loading...

Machine Learning-Based Building Climate Control Using Weather Forecast Data

Khakzad Gharamaleki, Sepideh | 2021

386 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 54469 (05)
  4. University: Sharif University of Technology
  5. Department: Electrical Engineering
  6. Advisor(s): Rezaeizadeh, Amin
  7. Abstract:
  8. Heating, Ventilation and Air-Conditioning (HVAC) systems of buildings consume an excessive amount of energy and emit even more amounts of carbon, all around the world. Rule-Based Control (RBC) algorithms, which switch the facilities on and off according to the measurements of the building’s sensors, are the most frequently utilized controllers in HVAC systems. Due to the conservative settings of the comfort-zone in RBC strategies, energy consumption increases by a large amount. One of the most conventional ways to improve the energy efficiency along with providing the thermal comfort of the occupants of the building, is model predictive control (MPC) algorithms. In order for MPC to work properly, a highly accurate model, large datasets and solving an optimization problem online is needed. In this paper, a precise model is developed to determine the exact loads and demands of a HVAC for MPC, by presenting a model both for the building’s heat system and the occupants’ temperature comfort-zone utilizing neural networks and in order to appraise the proposed strategy, it’s been compared to RBC techniques
  9. Keywords:
  10. Machine Learning ; Model Predictive Control ; Space Heating System ; Multizone Heating, Ventilation and Air Conditioning (HVAC) Systems Control ; Building Thermal Modeling ; Building Air Conditioning Controller System

 Digital Object List

 Bookmark

No TOC