Research Of Building Climate Classifications Based On Cluster Analysis And Energy-Saving Control

Specific climatic conditions are the dominant factors in building formation and evolution,and different climatic conditions have different architectural forms.Before the construction of relevant standards and policy development,the climate classifications are the most basic work.How to find valuable knowledge from the data and establish the modle of climate classifications are important to the energy-saving.The main aspects of the paper are as follows :(1)Using the comprehensive factor method to select the average temperature in January,the average temperature in July and July relative humidity,precipitation,the days less than temperature 5 ? and more than 25 ? as the classification index,and discussing the index primary zoning.The corresponding index method is used to determine the secondary classification index by spatial variance coefficient and correlation coefficient.(2)Selecting the k-medoids clustering algorithm to classify the climate,and the clustering internal validity index is used to determine the number.Considering the difference of the contribution rate of each meteorological element,the weighted kmedoids algorithm is used to classify the building climate in China.Then it interpolates by using the Kriging interpolation method,and research the applicability of passive design strategies about building.(3)Taking the residential building under the summer cooling condition in Xi'an as an example.It is discussed when the human comfort is widened by the passive design of the natural ventilation,then it could solve how much extent of the thermal comfort problem.This paper systematical y establishes the index system of building climate zoning,and 612 stations are completed in China.Then it is compared with the original building climate zones.On the basis of the first-level zoning,the second-level zoning is subdivided,and the secondary zoning index is selected rationally through the statistical method,which can reflect the regional difference and avoid the influence of the subjective factors in the zoning,making the zoning more reasonable and scientific.And taking the residential building in Xi'an as an example,through the joint analysis of TRNSYS and MATLAB simulation in the use of passive technology and model predictive control to improve the human thermal comfort temperature.