Analysis And Optimization Of Community Energy Consumption Based On Deep Learning

With the rapid development of China’s social economy and the continuous advancement of urbanization process,building energy consumption has been growing rapidly in recent years.As an important part of building energy consumption,the operation energy consumption of urban residential buildings has the characteristics of large volume and rapid growth.It is of great significance for the sustainable development of society to do this part of building energy conservation well.In the current information and intelligent background,building energy consumption real-time monitoring and load accurate prediction can provide optimization suggestions for the main energy consumption management,which can effectively improve the current situation of energy consumption on the demand side and promote the development of building energy conservation.Based on building operation and maintenance management,Internet of things,computer and other disciplines,taking community energy consumption management as the background,the main research work and contribution is as follows:Firstly,this paper summarizes the current situation of residential building energy consumption monitoring,analyzes the shortcomings of traditional building energy consumption monitoring,studies the advantages of BIM Technology Applied to residential building energy consumption monitoring,designs a set of residential building energy consumption monitoring system based on BIM technology based on the characteristics of community energy consumption and the demand of energy management,and elaborates its system components--perception layer,transmission layer and management layer.Then,the structure composition and network training methods of RNN and LSTM in deep learning are introduced in detail,and their advantages and disadvantages are analyzed.On this basis,a short-term load forecasting method for community electricity consumption based on GRU network is proposed,including data preprocessing,model building,model evaluation and other steps.The historical load data and environmental date are used to forecast the short-term load of the community.It is applied to the load data of a real community,and the data quantization rules are formulated according to the data characteristics in advance,and the data are screened by empirical mode decomposition and Pearson correlation coefficient,and the model parameters are determined by combining theoretical analysis and a large number of experiments,and good prediction results are achieved.Compared with the mainstream methods,the advantages of the proposed method are verified.Finally,the paper analyzes the role and advantages of the proposed energy consumption monitoring system and load forecasting method in the analysis and optimization of community energy consumption,and puts forward specific energy consumption management optimization suggestions for families,communities and cities.To sum up,this paper improves the community residential building energy consumption monitoring and community power load forecasting method,puts forward specific implementation suggestions for the analysis and optimization of community energy consumption,improves the community energy consumption management level,and provides effective technical support for the actual building energy conservation work.