摘要
为了更好实现建筑数据采集,提出一种基于模拟退火算法—粒子群优化算法—径向基函数(SA-PSO-RBF)修正算法的智慧建筑数据采集方法。所提出的方法以STM32F407ZGT6为主控芯片,选用温湿度传感器、PM2.5传感器、二氧化碳传感器采集建筑数据;利用箱型图和k均值聚类算法改进的支持向量数据描述(SVDD),分别对传感器采集的横向异常数据和纵向异常数据进行检测;利用SA-PSO对RBF神经网络参数进行优化,以用于对异常数据的修正,并利用四分位数概念对修正后的数据进行融合;通过Wi-Fi通信将修正后的数据上传到云平台存储,从而实现智慧建筑的监测。测试结果表明,所构建的系统采用SA-PSO-RBF对智慧建筑温湿度数据、PM2.5数据、二氧化碳数据进行修正,提升了数据的真实性。所构建的系统可实现智慧建筑数据的采集、传输、分类存储与查看,且具有低延时特点,能实时反映智慧建筑状态。
To better achieve building data collection,a smart building data collection method based on SA-PSO-RBF correction algorithm is proposed.The proposed method takes STM32F407ZGT6 as the main control chip,and selects temperature and humidity sensors,PM 2.5 sensors and carbon dioxide sensors to collect building data.The box plots and the improved support vector data description(SVDD)based on k-means clustering algorithm are used to detect lateral and vertical abnormal data collected by the sensors.Parameters of RBF neural network are optimized by SA-PSO for correcting abnormal data,and the corrected data are fused by the concept of quartile.The corrected data are uploaded to the cloud platform for storage through Wi-Fi communication,thereby achieving the monitoring of smart buildings.The test results show that the constructed system uses SA-PSO-RBF to correct the temperature and humidity data,PM 2.5 data and carbon dioxide data of smart buildings,improving the authenticity of the data.The constructed system can achieve the collection,transmission,classified storage and viewing of smart building data,and has low latency characteristics,which can reflect the status of smart buildings in real time.
作者
林再国
胡卿
LIN Zaiguo;HU Qing(Power China Huadong Engineering Corporation Limited(HDEC),Hangzhou 311122,China)
出处
《微型电脑应用》
2026年第1期48-53,共6页
Microcomputer Applications
基金
中国电建华东勘测设计研究院科技项目(KY2022-JZ-02-17)。
