Research On Energy Management Platform Based On Data Mining

In recent years,all countries in the world regard building energy conservation and emission reduction as a core content of economic development.China’s building energy consumption accounts for about 33% of the total social energy consumption,with the total construction and living cleanliness and comfort requirements continue to increase,building energy consumption will continue to show rapid upward trend,so the problem of building energy conservation is the most important of China’s energy-saving emission reduction.At present,China has carried out a series of work on building energy conservation and emission reduction,many scientific research workers have also done a lot of work in building energy consumption monitoring,energy consumption and energy saving analysis and fault diagnosis.This paper will use data mining technology,extracting data from the variables of building energy management platform system to build a building energy modeling,and a comprehensive office building as the research object,using the fault diagnosis model of building energy consumption,energy consumption of buildings in the running process of the abnormal disturbance identification and diagnosis research,and notify the maintenance personnel of abnormal events and system energy consumption the fault location,finallyto reach the purpose of fault elimination.Building energy consumption monitoring and diagnosis system is of great practical significance to improve the efficiency of energy use,to ensure the operation process,to ensure the safety of equipment and personnel.First,Principal component analysis is chosen as the method of data mining in this paper,PCA is the most widely used method in multivariate statistical analysis,to determine the number of principal components selection,determination of principal component analysis to establish rules model,determine the calculation method of the statistics and the control limit of the fault diagnosis based on the principal component analysis,realization of principal component analysis based on MATLAB program;Secondly,based on Skyspark software to establish building energy monitoring platform,all energy consuming equipment in building is integrated in the monitoring platform,and establish a meteorological station at the top of the building,collection of outdoor temperature and outdoor wind speed,humidity and outdoor PM_2.5 concentration.On this basis,according to the type and use of energy,statistics and monitoring of building energy consumption,selection of cooling consumption,water consumption,air conditioning electricity consumption,lighting electricity consumption,landscape electricity consumption,power consumption,living electricity consumption,production of electricity and commercial electricityconsumption,further analysis of the use of energy consumption,and the realization of energy consumption statistics and energy saving analysis,improve the wisdom of energy management platform;Thirdly,this paper selects the data of 100 consecutive days of building energy consumption input variables,and uses the MATLAB software to establish the fault diagnosis principal component model of building energy consumption system.Based on the actual data and principal component analysis,when the cumulative contribution rate of CPV（k）=87.046%,the number of principal components k=7,confidence alpha =99%,UCL=21.0524,Q=2.4262,to establish the model of intelligent energy consumption system,the diagnostic model in this case is identical to the actual process;Finally,through the energy monitoring platform collected 365 days a year of building energy consumption and operating data,application and test of the established intelligent building energy consumption system fault diagnosis model,and establish the corresponding rules of fault diagnosis,it is found that there are 11 UCL=21.0524 control charts beyond the control limit of T2 in the process of system operation,SPE statistical monitoring chart in the process of running the system beyond its control limit of Q=2.4262,the system will alarm,so as to judge fault reason.Using the above research achievements,when the energy consumption of the building system fault occurs,matching the system change characteristics with the principal component model,combined with the matching results of both can achieve the fault detection and diagnosis.The results lay a good foundation for the fault detection and diagnosis of building energy management system in the future.