A forecasting strategy based on wireless sensing for thermal comfort optimization in smart buildings

The prediction of thermal comfort in smart buildings is addressed in this article to support the energy manager in the management of the heating, ventilation, and air conditioning systems. An efficient control is fundamental to meet the user satisfaction as well as the energy cost reduction. The proposed method based on a customized support vector regression technique predicts and suggests to the energy managers the indoor target temperature required to obtain the desired comfort throughout the building. A network of low-cost wireless sensors has been deployed to measure the indoor conditions of a real indoor scenario. The prediction performance has been experimentally validated and the obtained results have pointed out a temperature prediction error lower than 18C and the achievement of the thermal comfort requirements.