基于机器学习算法的建筑物健康状态识别  

Building Health State Recognition Based on Machine Learning Algorithm

作　　者：张培培[1] 南江萍 王昭[1] ZHANG Peipei;NAN Jiangping;WANG Zhao(ZTE Communication Academy,Xi’an Traffic Engineering Institute,Xi’an 710300,China)

机构地区：[1]西安交通工程学院中兴通信学院,陕西西安710300

出　　处：《微型电脑应用》2020年第10期10-12,共3页Microcomputer Applications

基　　金：陕西省教育厅科研计划项目(18JK1042)。

摘　　要：为了提高建筑物健康状态识别结果,提出一种基于机器学习算法的建筑物健康状态检测方法。首先分析建筑物健康状态的识别流程,找到影响建筑物健康状态识别效果的因素,然后从中选择主要的影响因素进行建筑物健康状态识别建模,并引入机器学习算法描述建筑物健康状态与影响因素之间的内在联系,建立建筑物健康状态识别模型,最后采用具体建筑物健康状态识别实例分析了该方法的有效性和优越性,对建筑物健康状态识别率平均值超过92%,而当前经典方法的建筑物健康状态识别率没有超过90%,且识别速度更快,具有更好的实际应用价值。In order to improve the results of building health state recognition,a method of building health state detection based on machine learning algorithm is proposed.Firstly,the recognition process of building health status is introduced to find out the factors that affect the recognition effect of building health status.Then,several main factors are selected from the influencing factors of building health status to medel the recognition of building health status.Then,machine learning algorithm is introduced to describe the memory relationship between building health status and influencing factors,and the building health status is established.Finally,the effectiveness and superiority of the method are analyzed by a specific example of building health state recognition.The average recognition rate of building health state of the method preposed in this paper is more than 92%,while the recognition rate of building health state of the current classic method is less than 90%,and the recognition speed of building health state is faster,which has better practical application value.

关 键 词：建筑物 健康状态类型 机器学习算法 影响因素 应用实例 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]