Avaliação da eficiência energética de um sistema IoT de controle adaptativo para ar-condicionado baseado em conforto térmico

Abstract

This work presents the evolution and experimental validation of an IoT-based adaptive air- conditioning control system grounded in Fanger's thermal comfort model (ISO 7730). The prototype integrates BME280 sensors (air temperature and relative humidity) and an AMG8833 sensor (infrared thermal) to estimate the mean radiant temperature, in accordance with ISO 7726, aiming to overcome methodological limitations observed in previous studies, such as the lack of physical thermal radiation measurements and the limited accuracy of thermo-hygrometric sensors. A total of 20 field experiments were conducted (10 without the prototype, manual operation with variable setpoints, and 10 using the adaptive control system) in a standard office room. The achieved energy efficiency was ΔE = 54.33% (E_A = 46.29 kWh; E_B = 21.14 kWh), with a statistically significant difference (p < 0.002). The results validate the hypothesis that predictive control based on a thermo-physiological model can simultaneously provide energy savings and thermal comfort, while remaining consistent in order of magnitude with comparable studies reported in the Brazilian literature. Index Terms—Thermal comfort, ISO 7730, PMV, PPD, energy efficiency, ESP32, adaptive control, Internet of Things (IoT), air conditioning.