Energy Saving Of Urban Buildings Based On 3D Geographic Information System

With the rapid growth of urban scale in China,building energy consumption in urban environment has increased dramatically.Moreover,building energy consumption is one of the major causes for the haze problem in North China.Therefore,the city-scale building energy consumption model is necessary to fully analyze the characteristics of building energy consumption in urban environment,which has practical significance in promoting the reduction of both urban building energy use and haze problem,and also improving the quality of urban life.However,the simulation research of urban-scale building energy consumption in our country is still in a very elementary stage.There are many problems and shortcomings in massive energy use information processing,large-scale building modeling,and visual evaluation of energy consumption at urban scale.A systematical method of energy simulation and saving potential for urban building is proposed in this paper by combining building energy simulation and data processing technology based on the geographic information system(GIS)platform.A university campus located in Tianjin University of Science and Technology in HeXi District,Tianjin,is selected as a case study to demonstrate the method proposed for energy analysis of urban buildings.This research method can be divided into four steps,including geographic information system data processing,automated modeling method,machine learning-based building energy models,and energy saving visualization.First,the data processing of geographic information system puts forward the basic frame of building database for building energy consumption model based on GIS.Second,the automation of building city energy consumption model is realized by using 3D geographic information system.The computing environment used in this project is R statistical environment to automatically create the energy models required for a specific urban area.The energy model constructed can have five perimeter zones and one core zone on each floor of a building.Third,we compare the predictive performance of five machine learning methods and establish the optimal energy consumption statistical model for this case study.The computation speed could be increased by at least 102 multiples.Finally,the visualization method of energy use for urban buildings is proposed to display the spatial distribution of energy consumption by combining the machine learning energy meta-model,global sensitivity analysis,and GIS platform.Global sensitivity analysis can determine the importance ranking for different energy saving measures to create high efficient and flexible energy saving plan.The energy saving results can be shown in GIS platform by combing the fast computing energy models from machine learning algorithms.This study can promote the wide application of energy analysis models at the preliminary stage of urban energy planning to predict the energy performance of various energy saving schemes.Furthermore,the method proposed in this project will promote the development of smart city,and help the development of both building information model and urban planning.