Fig. 4: Global prediction of thermo-responsive (TR) window performance. | Nature Communications

Fig. 4: Global prediction of thermo-responsive (TR) window performance.

From: Global and regional perspectives on optimizing thermo-responsive dynamic windows for energy-efficient buildings

Fig. 4

a–d Training and testing results of artificial neural network (ANN) models for global vertical irradiance (GVI) (a), total site energy saving (\(\Delta {E}_{{{\rm {TR}}}}\)) (b), necessity level (\(\Delta {E}_{{\rm {n}}}\)) (c) (b and c are trained by data under optimal conditions), and optimal transition temperature (opt \({T}_{{{\rm {tran}}}}\)) (d) (d is trained by data under all conditions). e–h World heatmaps of \(\Delta {E}_{{{\rm {TR}}}}\) (e), \(\Delta {E}_{{\rm {n}}}\) (f), TRRI (g) (under optimal conditions), and opt \({T}_{{{\rm {tran}}}}\) (h) (when \({\tau }_{{{\rm {clear}}}}=0.8\), and \({\tau }_{{{\rm {dark}}}}=0\)). Coastlines, country, and state boundaries are provided by the Natural Earth dataset (public domain). Source data are provided as a Source Data file.

Back to article page