Study On Energy Consumption And Comfort Of Residential Building Based On Genetic Algorithm And Data Mining

With the rapid development of the economy,more and more attentions have been paid to the comfortable living environment.In general,improving indoor thermal comfort level will also lead to increased energy consumption in residential buildings,and the need to find the dynamic balance between building energy efficiency and indoor thermal comfort has become the focus.According to different weather conditions,five major climatic regions have been divided in China.The different building energy efficiency design standards/regulations have been developed,which can only applied to the local climatic condition.This research try to seek common design methods to overcome the limitation of the design standards/regulations so that it can be applied to different climate regions,by integration of building simulation software and genetic algorithms.NSGA-? is the most widely used multi-objective optimization algorithm.It optimizes the window-to-wall ratio for each facade,the building orientation,external walls/roof of the building construction,cooling/heating temperature setpoints,and HVAC system types.An optimized database considering both building energy consumption and indoor discomfort degree hour in typical cities in the five climatic regions was developed.Through the analysis of the optimal solution,there is big energy saving in the northern region with maximum reduction of 38.37%,and the number of discomfort degree hour can be reduced by up to 61.5%.Finally,a representative data set was built to apply data miningtechniques.The accuracy of seven different data mining algorithms was compared and analyzed to find a data model that can be used to quickly predict the energy consumption and discomfort degree hour.The results show that the BPNN model has the highest accuracy and best adaptability.The prediction accuracy of building energy consumption and indoor thermal comfort reach 98.14% and 97.72%,respectively.