基于机器学习算法的高层建筑电气设备分类  

Classification of Electrical Equipment in High-rise Buildings Based on Machine Learning Algorithms

作　　者：杨光 陈健 付宝鑫 Yang Guang;Chen Jian;Fu Baoxin(Changzhou Sino Sea Elecpower Co.,Ltd.,Changzhou,Jiangsu 213000,China;Tianjin Research Institute of Electric Science Co.,Ltd.,Tianjin 300180,China)

机构地区：[1]常州中海电力科技有限公司,江苏常州213000 [2]天津电气科学研究院有限公司,天津300180

出　　处：《机电工程技术》2024年第7期231-234,314,共5页Mechanical & Electrical Engineering Technology

摘　　要：高层建筑内电气设备易因电能质量暂态扰动而造成恶劣影响,需对电气设备进行扰动耐受特性测试来确定待保护的设备,而在繁多的电气设备中,如何快速高效选择待测试电气设备成为亟待解决的问题。为此,通过研究朴素贝叶斯算法在电气设备分类模型上的应用,考虑高层建筑电气设备类型,将电气设备根据5种与扰动耐受特性相关的特征划归为4种类别,基于先验信息和有标签的电气设备构建了面向高层建筑的电气设备分类模型。采用某高层建筑的多类电气设备作为数据集对分类模型进行训练与测试,并探讨了该分类模型的准确性和有效性。分类结果显示,所提出的基于朴素贝叶斯算法的高层建筑电气设备分类模型在该数据集上具有较好的分类效果,为高层建筑电气设备快速分类提供了新的研究方向和思路。Electrical equipment within high-rise buildings is vulnerable to detrimental consequences resulting from transient power quality disturbances.To identify equipment necessitating protection,it becomes imperative to conduct disturbance tolerance testing.However,the task of efficiently and effectively selecting the appropriate equipment for testing from a wide array of options presents a challenge.This study aims to address this predicament by investigating the implementation of the Naive Bayes algorithm in an electrical equipment classification model specifically tailored to high-rise buildings.The classification process involves categorizing the electrical equipment into four distinct classes based on five features associated with disturbance tolerance.By capitalizing on prior knowledge and utilizing labeled data pertaining to electrical equipment,a classification model is developed specifically for high-rise buildings.The model is subsequently trained and tested using a dataset comprising diverse types of electrical equipment obtained from a specific high-rise building.The accuracy and effectiveness of the classification model are thoroughly evaluated.The findings demonstrate that the proposed Naive Bayes-based electrical equipment classification model achieves commendable performance on the dataset,thereby introducing a fresh research direction and approach for expedited classification of electrical equipment in high-rise buildings.

关 键 词：机器学习 高层建筑 电气设备 

分 类 号：TU976.1[建筑科学]