| Building operation energy consumption is an important part of the energy consumption of the whole life cycle of a building,and it accounts for a large proportion of the total energy consumption of the society.Therefore,it is possible to effectively simulate and predict building energy consumption in the urban planning and building design phases as well as in the later operation and maintenance phases,which can provide positive feedback on planning,design and operation.Existing building energy consumption simulation is a bottom-up simulation method based on building monomers.This method cannot be applied to city-level and regional-level building energy consumption planning.Aiming at the limitations of existing building energy simulation methods to deal with the energy demand of urban buildings,this paper proposes a data-driven model of urban building energy consumption assessment method,and analyzes the data-driven model from four aspects: technical background,framework foundation,system optimization and practical application.Research on the simulation and prediction method of building energy consumption.This paper sorts out the basic process of building energy consumption,summarizes the inherent contradictions between the existing building energy consumption assessment methods and the development trend of urban building energy demand,and then proposes a data-driven model of building energy consumption simulation assessment methods: including energy consumption data sets Four parts: construction and optimization,urban building automation energy modeling,machine learning algorithm application,and quantitative analysis of space energy consumption.First,establish a data-driven framework,methods and specific steps for building energy consumption data,and complete the preliminary data construction through Arc GIS and Geo Pandas.Secondly,the three core data nodes of climate environment,geographic information and building energy consumption were systematically analyzed,and targeted optimization methods were proposed to improve the accuracy and timeliness of the building energy consumption data set.Third,apply machine learning algorithms to building energy consumption simulation prediction and cluster analysis,specifically evaluate 9 classification algorithms,and select RF as the algorithm model for building type classification;on the basis of classification,evaluate 3 regression algorithms and select XGB is used as a regression algorithm for building energy consumption simulation;K-Means algorithm is used to analyze the distribution of building energy consumption in geographic space from a macro perspective.Finally,a method of spatial quantitative analysis is proposed to accurately describe and evaluate building energy consumption,and complete the data-driven model of building energy consumption evaluation from construction,optimization,machine learning energy consumption proxy model regression prediction,and energy consumption data visualization.Method verification of the process.The results show that,compared with traditional engineering simulation methods,the data-driven model of urban building energy consumption simulation and prediction proposed in this paper can realize rapid simulation and prediction of urban building energy consumption with small errors,effectively saving time And computing resources can provide a rapid overall energy consumption assessment reference for planning and design practices,and accelerate the development and improvement of urban energy systems. |
