Adaptive optimal design of active thermoelectric windows using surrogate modeling

Abstract

This paper develops an optimal design and an optimal operating strategy for Active Thermoelectric (ATE) windows. The proposed ATE window design uses thermostats to actively control thermoelectric (TE) units and fans to regulate the heat transfer through the windows. To achieve high energy efficiency, optimization of the ATE window design seeks to simultaneously minimize the heat transferred through the window and the net power consumption. The ATE windows should adapt to varying climatic conditions. The heat transfer and the power supplies are optimized under a prescribed set of climatic conditions. Based on the optimal results obtained for these conditions, surrogate models are developed to represent the optimal modes of operation as functions of the climatic conditions, namely (i) ambient temperature, (ii) wind speed, and (iii) solar radiation. To this end, Radial Basis Functions (RBF) are used. The results show that the ATE windows provide significantly improved insulation compared to traditional windows under varying climatic conditions. Moreover, it was found that the ATE window operates at a superior efficiency than a standard HVAC system, particularly in colder climates.