基于低耗能通风的建筑空间形态特征参数测试  

Test of Building Space Shape Characteristic Parameters Based on Low Energy Consumption Ventilation

作　　者：李雪松 LI Xue-song(School of Civil Architecture and Environment,Hubei University of Technology,Wuhan Hubei 430068,China)

机构地区：[1]湖北工业大学土木建筑与环境学院,湖北武汉430068

出　　处：《计算机仿真》2020年第6期369-373,共5页Computer Simulation

摘　　要：利用传统方法测试建筑物空间形态特征参数时,得出的结果不够准确,因此室内自行通风效果较差,多数依赖于空调系统通风,不仅造成环境污染而且不利于人们的身体健康。基于此提出低耗能通风的建筑空间形态特征参数测试。首先对低耗能通风的作用原理、主要途径和设计方法介绍;在此基础上,获取超分辨率SAR的建筑影像,并对影像做预处理,得到变异系数,利用滤波准则并确定不同层级图像得出表达式;最后通过人工神经网络算法实现从输入空间变换成输出空间的映射,从而完成建筑空间的形态特征参数测试。实验结果表明,所提方法测试出的参数准确性高,相对与其他方法操作简便,可以满足低耗能通风建筑的要求。When the traditional method is used to test the characteristic parameters of building space form,the results are not accurate enough,so the effect of indoor ventilation is poor.Therefore,this article puts forward a method to test the building space form feature parameters based on low energy consumption ventilation.Firstly,the principle,main ways and design method of low energy consumption ventilation were introduced.On this basis,the building image of super-resolution SAR was obtained,and the image was preprocessed to get the coefficient of variation.Secondly,the expression was derived by using the filtering criteria and determining the images in different levels.Finally,the mapping from the input space to the output space was achieved by artificial neural network algorithm.Thus,the test of morphology feature parameter was completed.Simulation results show that the parameters measured by the proposed method are extremely accurate.It is easy to operate,so this method can meet the requirements of ventilated buildings with low energy consumption.

关 键 词：低能耗通风 建筑空间 形态特征 神经网络 超分辨率图像 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]