Missing Data Filling Methods of Air-Conditioning Power Consumption for Public Buildings

At present, some public building data platforms are built for building operation data analysis. Data missing is a common problem in these platforms and it is necessary to fill in the missing data for correct data analysis. BP neural network algorithm is used to fill in the missing data of air-conditioning (AC) power consumption in public buildings. The correlation coefficient method is used to verify the correlation of AC power consumption with outdoor and indoor factors. Based on the correlation coefficient, the input parameters of the network model are determined. Taking the energy consumption of AC in a shopping mall in the refrigeration season as an example, the experimental results show that the BP neural network filling model after optimizing the input parameters can meet the filling accuracy requirements when the AC power consumption data is missing.