A TLBO-Tuned Neural Processor for Predicting Heating Load in Residential Buildings

Recent studies have witnessed remarkable merits of metaheuristic algorithms in optimization problems. Due to the significance early analysis thermal load energy-efficient buildings, this work introduces and compares four novel optimizer techniques—the firefly algorithm (FA), optics-inspired (OIO), shuffled complex evolution (SCE), teaching–learning-based (TLBO)—for an accurate prediction heating (HL). The models are applied a multilayer perceptron (MLP) neural network surmount its computational shortcomings. fed by literature-based dataset obtained for residential buildings. results revealed that all used capable properly analyzing predicting HL pattern. A comparison between them, however, showed TLBO-MLP with coefficients determination 0.9610 vs. 0.9438, 0.9373, 0.9556 (respectively, FA-MLP, OIO-MLP, SCE-MLP) root mean square error 2.1103 2.5456, 2.7099, 2.2774 presents most reliable approximation HL. It also surpassed several methods previous studies. Thus, developed can be beneficial model subsequent practical applications.