An energy consumption rectification method based on Bayesian linear regression and heating degree-days

The time-varying external environment is one of the main variables influencing heating energy consumptions, so that its influence should be rectified when energy savings of different heating modes are calculated. This paper proposes an energy consumption rectification method based on Bayesian linear regression and heating degree-days, to obtain heating energy consumptions without the influence of different outdoor temperatures. The proposed method consists of three main steps. First, a physical model of heating houses is used to prove a relationship between energy consumptions and heating degree-days. Second, Bayesian linear regression is exploited to estimate uncertainty ranges of heating energy consumptions. Finally, heating energy consumptions are rectified, and energy savings with their uncertainty ranges for different heating modes under the same outdoor temperature are obtained. The proposed method does not require the physical parameters of heating houses to facilitate practical implementation. Additionally, it provides uncertainty ranges of heating energy consumptions to measure the estimation accuracy. Numerical and experimental examples show that the proposed method provides more accurate estimates of heating energy consumptions than existing methods.