Heat Pump Energy Consumption Optimization Using Machine Learning

The current economic crisis in Europe, caused by the severe years of COVID and the invasion of Ukraine, has led to a significant energy crisis, especially in countries with lower temperatures. This energy situation has prompted a heightened emphasis on enhancing the efficiency of residential building systems, specifically Heating, Ventilation, and Air Conditioning (HVAC) systems. Additionally, the increasing demand for smart homes has created a need for intelligent HVAC control solutions. The objective of this work is to develop an intelligent control system using ML techniques to optimize costs in HVAC systems encompassing a Heat Pump (HP). The study aims to create a tool that generates a suitable Time Program (TP) based on observed operating conditions and user needs, a model for predicting HP consumption based on different operating temperatures, and an intelligent control of the HP heating curve to minimize operational costs. Results show that implementing the proposed ML-based TP optimization leads to significant energy consumption reduction in the HP, with an average reduction of about 16.3%. The algorithm takes into account various factors, such as adjustments to the SP temperature, to optimize energy efficiency based on outdoor temperatures and user preferences.