A novelty prediction approach of natural daylight in classroom using ANFIS model

Natural daylight plays an important role in energy consumption of buildings. Natural daylight passing through the building facades effects the working space inside the room. According to the LEEDv4 standard, in order to evaluate the ability of natural daylight inside the rooms, some technical quantities are commonly employed such as sDA, ASE, and DF. However, numerical simulation techniques usually take a lot of time and computer resources. In this study, we surveyed and evaluated different sizes of Box-framed louvres facades and propose a novelty prediction approach, named ANFIS which composes of the human experience and artificial intelligence. This method predicts the coefficients of natural daylight through the building façade affect. By using the Adaptive Neuro-Fuzzy Inference System technique (ANFIS), the result shows a great accuracy on prediction. Keywords: Natural Daylight, Daylight Factor, Adaptive Neuro-Fuzzy Inference System, Machine Learning