Premium
Building Heat Load Estimation Method Including Parameter Estimation from Actual Data
Author(s) -
Nakamura Ryosuke,
Kawamura Tsutomu
Publication year - 2021
Publication title -
ieej transactions on electrical and electronic engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.254
H-Index - 30
eISSN - 1931-4981
pISSN - 1931-4973
DOI - 10.1002/tee.23403
Subject(s) - kalman filter , control theory (sociology) , stability (learning theory) , estimation theory , air conditioning , thermal , power (physics) , extended kalman filter , computer science , engineering , algorithm , meteorology , mechanical engineering , control (management) , physics , quantum mechanics , machine learning , artificial intelligence
In recent years, various method of air conditioning control and making operation plans have been developed. They require a building thermal model, but for making the model detailed design data is needed. Several model construction methods have been proposed up to now, but they are for buildings surrounded by walls made with one material. In this situation for deriving the room temperature and air conditioner power consumption, a parameter estimation method for estimating heat capacities and resistances of walls with different physical properties was developed. This method consists of two steps. At the first step, in steady temperature state, the least‐squares method is used to derive thermal resistances. In the second step, in non‐steady temperature state, an Unscented Kalman Filter (UKF) is used to derive heat capacities and solar‐radiation shading coefficients. UKF is adopted for numerical stability and short operation time. With this method, it is possible to obtain the independent parameters of each wall by repeating the evaluation using a simple model. These steps were applied to data obtained by simulation. As a result, the physical parameter error of the inner and outer walls was sometimes 10–60%. On the other hand, regarding to the estimation target, the error of the indoor temperature estimation was 0.3 °C, and the error of power consumption estimation was less than 4%, so high estimation accuracy was obtained. This confirmed the effectiveness of the proposed method. © 2021 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC.