Public Buildings Can Be Used To Monitor Statistical Analysis Of City-level Platform Data

With the steady growth of economy and the progress of the society.China's energy consumption is increasing,the problem which is large-scale public buildings height energy consumption has become increasingly prominent.Response to this situation,this paper takes Shanghai as an example,selected the energy audit data of nearly 700 buildings and the energy consumption statistics of 1400 building and the structured and semi structured data of building energy consumption for all levels of government and society as the basis.Building statistical analysis models,in order to effectively promote energy conservation work,help the government to establish an effective large public building energy monitoring system.Firstly,Screening of effective data,the large public buildings in Shanghai are divided into ten categories.As the average monthly consumption of large public buildings and the average annual consumption are important indicator for government supervision departments evaluate the work of energy-saving emission reduction,has an important guiding significance.Secondly,this paper uses the EM algorithm to interpolate the missing data of energy consumption of large buildings,which effectively solves the problem of missing data in the statistical analysis of the energy consumption of large buildings in Shanghai.This paper through the establishment of a hierarchical Bayesian model and Gibbs sampling method to estimate monthly consumption and annual consumption of the various types of large buildings in Shanghai city in 2015.Because there is a correlation between the average consumption of each month,this paper defines the first-order autocorrelation structure to adjust the correlation.Finally,the comparison analysis found that the energy consumption per unit area of office buildings has a strong cyclical fluctuation,and due to the shopping malls are open 7 days a week,its consumption is not obviously periodic.Therefore,this paper use the two types of buildings as representatives,and we use time series to analysis energy consumption per unit area of the two representatives which are "Metro" and "New Horizons",established the model of consumption which is cyclically adjusted and temperature variables.The AR model is established to describe the correlation between sequences.