Research on building energy consumption model based on multiple regression

The benchmarking and evaluation of the energy consumption of public buildings can provide a data basis for building energy-saving work during the operation phase and is an important guarantee for the actual energy-saving effect. In this paper, the life cycle carbon emission of Chinese buildings is taken as the research object. Through scenario simulation, the evolution trend of future carbon emission of macro construction sector is deeply discussed, and the key stages and key factors affecting the peak of carbon emission are analyzed. In addition, the responsibility of carbon emission reduction is effectively and fairly shared based on the actual situation. The main results of this study can enrich the discussion on the current situation of life-cycle energy consumption and carbon emissions of the macro building sector to a certain extent, and also further improve and expand the empirical method and theoretical system of the driving factors of carbon emissions of the macro building sector and the prediction of carbon emissions peak. From a practical point of view, the results of this study will be helpful for the Chinese government to predict the future development of carbon emissions in the macro construction sector, and also provide reference for the future "14th Five-Year Plan" emission reduction target setting or more long-term energy conservation and emission reduction policy planning. Then, based on the complex and variable non-steady-state characteristics of the building Mechanical Electrical & Plumbing system, a Non-Intrusive Load Measurement method using composite classification algorithm is established for the equipment energy splitting, and the accuracy of which is obtained by the application in an example project is significant improved.