Design and Modeling of Intelligent Building Office and Thermal Comfort Based on Probabilistic Neural Network

Thermal comfort is strictly related to the efficient use of the environmental and energy resources to maintain or improve the quality life and thermal well-being of the residents. This article proposes to integrate the architectural building information and processing real-time data collected by IoT system to design room office. The control of room climate, lighting, and sun protection is the major goal of this research activity, which has the advantages of simplifying user operation, improving the energy efficiency, and ensuring the minimum energy consumption. Temperature, relative humidity, pressure and lighting sensors have been installed in office room. For the simulation of a virtual working room, a 3D model has been employed. The Spatial Daylight Autonomy (sDA), and Annual Sun Exposure (ASE), the annual glare distributions and lighting energy demand have been calculated in the simulated room. The thermal comfort in the building room is controlled by a digital sensors system connected to micro-controller module. Therefore, a neural network method is proposed to enhance the data collection and classification efficiency for an advanced thermal comfort analysis in the building office. The proposed method was validated by means of a confusion matrix exhibiting the correct classifications of 100% of samples within the testing dataset.