Research And Application Of Building Energy Consumption Early Warning Technology

Building energy conservation was one of the key areas of energy saving and emission reduction in China.In order to achieve energy consumption monitoring,building energy consumption monitoring system was builded,and energy consumption data was collected and upload by metering devices.Building energy consumption monitoring system can find the abnormal state of building energy in time with consumption warning function.Base on the analysis and summary of related litera,advantages and disadvantages of different prediction method about building energy consumption were analyzed.The development process and problems of building energy consumption monitoring system were discussed.Finally,it was pointed out that the traditional energy consumption warning method is low precision.First,the building energy consumption statistics and analysis of 4 universities in Fujia was done.The energy consumption of 6 other universities were also introduced.In this 10 universities,the average energy consumption per unit area is 6.59 kgce/(m2a),the energy consumption of per person is 149.84 kgce/(pa).Energy consumption in different universities was discussed.The result shows that the total construction area of campus,the number of students and the level of scientific research influence the energy consumption in colleges.The working principle,basic architecture,hardware composition,software function of the platform was expounded.The operation of a platform was introduced.The platform were monitoring 48 campus buildings.Aiming at the problems of low precision and poor adaptability in building energy consumption alarm threshold setting,the Elman neural network was used to establish the energy consumption prediction model,to calculate the energy consumption threshold.the principal component analysis(PCA)is used to the data simple in order to remove the redundant information and to de-correlate the variables.Furthermore,the principal components is inputted Elman neural network which is extracted by PCA.Finally a building energy consumption prediction model is build based on PCA-Elman.The PCA-Elman is applied to the energy consumption for a building,and the relative errors are 5.49%respectively.This show that the PCA-Elman is effective for building energy consumption prediction.This method can be used in building energy consumption prediction.It is the relation and law between things which the association rules are used in.The association rules and its evaluation indexes were discussed in detail.The Apriori algorithm is used in early warning of a university in Fujian.The results show that the conclusion obtained from the correlation analysis is similar when the structures of the building and the states of energy consumption are similar.The platform is provided with branch energy consumption alarm,total energy consumption alarm of a single building and alarm of equipment operating status.The PCA-Elman model is applied to the total energy consumption of a single building.It can be achieved on the platform by MATLAB and.NET mixed programming.The application results show that the method can improve the accuracy of building energy consumption threshold.