云计算下基于物联网的机械产品装配信息远程监测技术  

Remote monitoring technology for assembly information of mechanical products based on Internet of Things under cloud computing

作　　者：戚凌钧[1] Qi Lingjun(Zhejiang Technical Institute of Economics, Zhejiang Hangzhou, 311100, China)

机构地区：[1]浙江经济职业技术学院,浙江杭州311100

出　　处：《机械设计与制造工程》2019年第3期55-58,共4页Machine Design and Manufacturing Engineering

摘　　要：针对目前远程监测技术存在误差大、监测时间难以满足实际需求等问题,提出一种基于物联网的机械产品装配信息远程监测技术。在云计算环境下,对机械产品装配信息进行主成分分解,并在装配数据主成分不同区域建立不同子块,将原始装配数据特征空间划分为多个子空间,对不同特征空间内的装配数据子块提取其相关变化特征;根据装配特征相关变化特征,利用K-均值聚类算法对样本数据进行聚类,在机械产品各装配阶段构建代表性统计分析模型,采用该模型对整个装配过程进行远程监测。实验结果表明,该技术与当前远程监测技术相比,漏告警率可减少到0,误告警率降低了2%,且在实际环境下远程监测时间能够满足实际需求。Aiming at the problems of the current remote monitoring technology error and the difficulty of meeting the real-time needs in the actual environment, a remote monitoring technology for mechanical product assembly information is proposed based on the Internet of Things. In the cloud computing environment, the principal component decomposition of the mechanical product assembly information is performed, and different sub-blocks are established in different directions of the assembly data principal components, the original assembly data feature space is divided into multiple sub-spaces, and the assembly data in different feature spaces is assembled. The sub-block extracts its related variation characteristics;K- means clustering algorithm is used to cluster the sample data according to the relevant feature of the assembly feature, and a representative statistical analysis model is built in each assembly stage of the mechanical product, and the model is used to remotely monitor the entire assembly process. The experimental results show that compared with the current remote monitoring technology, the remote monitoring technology based on the Internet of Things reduces the false alarm rate to zero, and the false alarm rate is reduced by 2 percentage points. In the actual environment, the remote monitoring time can meet the real-time needs.

关 键 词：云计算 物联网 机械产品 装配信息 远程监测 主成分分析 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]