基于深度学习算法的物联网资源信息自动采集方法  

Automatic Collection Method of Internet of Things Resource Information Based on Deep Learning

作　　者：严梅 YAN Mei(ZheJiang Institute of Mechanical&Electrical Engineering,Hangzhou 310053 China)

机构地区：[1]浙江机电职业技术学院,浙江杭州310053

出　　处：《自动化技术与应用》2025年第2期61-65,共5页Techniques of Automation and Applications

基　　金：浙江机电职业技术学院教育教学改革项目(S-0152-21-471);浙江机电职业技术学院项目(S-1001-21-300-04-110)。

摘　　要：常规物联网资源信息自动采集方法,多采用本体构建技术,无法对其中的冗余信息进行有效滤除,导致资源信息采集效率较低。因此,提出基于深度学习算法的物联网资源信息自动采集方法。根据物联网运行特点设定物联网资源信息采集周期,控制采集过程,利用卡尔曼滤波法清洗原始数据,滤除冗余信息,提取关键信息进行内部融合,基于信息融合结果对其特征向量进行提取,分析信息概率分布情况,利用深度学习算法扩展信息采集量的最大化范围,进而实现物联网资源信息自动采集。采用仿真实验的方式验证所提方法的应用性能,结果证明,所提方法具有更高的资源信息采集效率。Conventional IoT resource information automatic collection methods mostly use ontology construction technology,which can not effectively filter redundant information,resulting in low resource information collection efficiency.Therefore,an automatic col-lection method of IoT resource information based on deep learning algorithm is proposed.According to the operation characteris-tics of the Internet of Things,it sets the collection cycle of the resource information of the Internet of Things,controls the collec-tion process,uses Kalman filtering to clean the original data,filters out redundant information,extracts key information for inter-nal fusion,extracts its feature vector based on the information fusion results,analyze the probability distribution of information,uses the depth learning algorithm to expand the maximum range of information collection,and then realizes the automatic collec-tion of the resource information of the Internet of Things.The application performance of the proposed method is verified by sim-ulation experiments,and the results show that the proposed method has higher resource information collection efficiency.

关 键 词：深度学习算法 物联网 资源信息 自动采集 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]