Performance analysis of space heating smart control models for energy and control effectiveness in five different climate zones

Abstract This paper compares smart control models for heating supply air among five different climate conditions to discuss the effectiveness of machine learning tools in terms of control and energy efficiency. A thermostat on/off control is typically used to maintain room temperature at a desired level. Advanced computing technologies have recently been introduced to complement the conventional on/off controls to improve control efficiency in heating systems. However, these methods, which were mostly utilized to control fuel amount or fan motor speed, lacked the capability to promptly respond to various outdoor temperature conditions as climate zones requiring refined control strategies to reduce environmental impacts. This paper proposes intelligent controls of mass and temperature simultaneously for heating air supply. The Fuzzy Inference System (FIS) and Artificial Neural Network (ANN) algorithms are utilized to develop six control models, and the models are tested to evaluate both control and energy efficiency during the winter season in five climate zones (from climate zone 2 through 6; i.e., Houston, Dallas, Raleigh, Chicago, and Detroit, respectively). Results include the energy consumption, control errors, and control signals in comparison to the baseline on/off control, which confirms the fact that the ANN simultaneous controls of mass and temperature is more effective than the other controllers for control accuracy and energy savings by 71.3% and 0.3%, respectively. The effectiveness of the ANN controller can contribute to maintaining room temperature accompanying the reduction of energy consumption, which is directly related to improve human comfort and reduce environmental impacts in various climate zones.