Studies On Fractal Characteristics And Forecasting Model Of Energy Consumption Of A Large Public Building

More than 25% of the energy consumption of our national buildings is consumed by large public buildings,but its energy efficiency is only 30%.The characteristics of large public buildings,which is "high energy consumption and low energy efficiency",is reflected.Building energy consumption prediction is an indispensable part of building energy management and building energy efficiency.The energy consumption of large public buildings has multi-variable,nonlinear,strong coupling and multi-perturbation complex characteristics.Analysis of energy consumption characteristics is the key to establish energy consumption prediction model.At present,most prediction algorithms fail to fully consider the complex characteristics of building energy consumption,resulting in unpredictable prediction results.Fractal theory can directly find some laws of its own changes from abstract unsimplified complex nonlinear things,and then analyze and predict it correctly.Therefore,it is a new way to apply fractal theory to analyze and solve the problem of the energy consumption of large public buildings.The thesis takes a commercial complex building in Xi’an as the research object.The research contents are as follows:(1)The fractal characteristics of the building energy consumption are analyzed from both qualitative and quantitative aspects,indicating that the building energy consumption has fractal characteristics,which is approximate self-similarity.(2)According to the influence of meteorological conditions,day type,date gap and other factors on building energy consumption prediction,and quantify the above factors,using fuzzy clustering theory,the fuzzy similarity matrix is established,and thesimilarity between each historical day and forecast date is calculated.Degree,selected the appropriate similar day.Among them,in view of the characteristics of the energy use of major holidays is different from the general use of energy,the principle of selecting similar dates for major holidays is proposed,and the rationality is verified by experiments.(3)Based on the similar day,the fractal interpolation algorithm is used to establish the energy consumption prediction model.Also,the results of the previous model and the BP neural network energy consumption prediction model are compared.The average relative error MRE and the root mean square error RMSE are used to evaluate the performance index.The experimental results show that the MRE value of the fractal prediction model is only 2.81%,and the RMSE value is also 1/6 of the RMSE value of the BP neural network prediction model.(4)Developed the energy consumption fractal prediction module of the large public buildings energy management information system,realized the function of energy consumption prediction,and designed the management function interface.The fractal characteristics of a commercial complex building in Xi’an are analyzed,and the fractal prediction model of energy consumption is established.The function of energy consumption prediction module is designed and developed,which can provide scientific basis for energy management and energy saving control of such buildings.It is conducive to timely grasping the energy trend of buildings and rationally distributing energy,laying a good foundation for the next step of energy conservation planning and energy management.