Power Consumption Analysis with Independent Component Analysis

Power Consumption Modeling of Airconditioning Units in an Educational Building Using Polynomial and Neural Network Fitting Techniques

“…Equation 7 describes how the power consumption, P(T,H), of an ACU is influenced by the levels of dry bulb temperature, T, and relative humidity, H. The polynomial curve fit coefficient values (k1 to k5) differ for instructional rooms and office spaces. P(T,H) = k1 + k2T + k3H + k4HT + k5H 2 (7) For instructional rooms: k1 = 2.9919, k2 = -0.0924, k3 = -0.0512, k4 = 0.0014, k5 = 0.0010. For office spaces: k1 = 1.7940, k2 = -0.0442, k3 = -0.0463, k4 = 0.0009, k5 = 0.0002.…”

“…Related researches include [3] using dynamic and static and hybrid data analysis in buildings, Related research such as [4] using K-means and [5] using kshape and random forest (KS-RF), to classify users according to power consumption behavior based on grid demand. A seasonal approach of educational building consumption from daily usage was limited to descriptive analysis through data cleaning and visualization [6], while [7] used Independent Component Analysis (ICA) to determine factors affecting consumption. A comprehensive review of building energy consumption prediction employing various neural network and regression methods, published between 2015 to 2022, was conducted by Borowski and Zwolińska [8].…”