Data Driven Thermal Comfort Model for Smart Home Energy Management System

Recent years have seen an increase in the popularity of smart and energy-efficient homes. The main issue in developing a control system for such a structure is to reduce energy usage without sacrificing customer satisfaction. This article has proposed a multi-objective paradigm to achieve this goal. Visual, air quality, and thermal comfort are considered. Particle swarm optimization (PSO) is utilised to optimise the overall system. It is still difficult to ensure that all occupants are satisfied with their thermal comfort because of how various people's body temperatures are established. In this paper, a data-driven approach is proposed to predict user thermal comfort in residential houses. The interior comfort temperature of each individual occupant has been predicted using an artificial neural network (ANN) prediction model, which may be utilised as the comfort temperature reference for heating, ventilation, and air conditioning (HVAC) management systems. The proposed model has been compared with single objective comfort maximisation with variable thermal comfort set points for individual occupants and a constant set point for all occupants, and the proposed model has shown its efficacy.