一种新的IoT PaaS大数据服务平台的设计  

Design of a New IoT PaaS Big Data Serving Platform

作　　者：孙丽娜[1] 武海燕[2] SUN Lina;WU Haiyan(Henan University Minsheng College,Kaifeng 475004,China;Image and Network Investigation Department,Railway Police College,Zhengzhou 450053,China)

机构地区：[1]河南大学民生学院,河南开封475004 [2]铁道警察学院图像与网络侦查系,郑州450053

出　　处：《重庆理工大学学报（自然科学）》2019年第7期160-166,共7页Journal of Chongqing University of Technology：Natural Science

基　　金：国家自然科学基金面上项目(11671119);河南省科技攻关项目(182102210118);河南省高等学校重点科研项目(18A520045,19B520026);中央高校基本科研项目(2017TJJBKY017,2018TJJBKY017);公安部技术研究计划项目(YJ20187283)

摘　　要：大数据时代,工业设备传感器会不间断产生海量数据,大数据储存与处理的效率成为服务平台研发主要关注的问题。为了解决上述问题,提出一种新的大数据服务平台,即物联网平台服务(Internet of Thing platform as a service,IoT PaaS)。本平台由注册模块、数据收集模块与分布式运算模块组成,在Mesos与Spark分布式架构中实现。另外,实现了RESTful API风格的平台通信接口的设计。该平台能够对来自产业领域中机器或设备传感器产生的大数据进行处理与分析服务,可快速实现大量数据在输入分析模型前的预处理,对于探索提升智能产线效能因素具有很好的决策参考意义。采用计算机计算能力、数据写入处理时间及大量数据查询时间作为平台运作量测指标。实验结果表明:该平台能够有效地为客户端解决终端设备大量数据处理与分析的瓶颈问题。In the era of big data,industrial equipment sensors will generate massive amounts of data without interruption.The efficiency of big data storage and processing has become a major concern of R&D service platforms.In order to solve the above problems,this paper proposes a new big data service platform,namely Internet of Thing platform as a service (IoT PaaS),which is composed of registration module,data collection module and distributed computing module,and is implemented in the distributed architecture of Mesos and Spark.In addition,the design of the platform communication interface of the RESTful API style is implemented.The platform of this paper can provide processing and analysis services for big data generated by sensors from machines or equipment in the industrial field.It can quickly realize the pre-processing of large amounts of data before inputting the analysis model.It has a good decision-making reference significance for exploring factors to improve the efficiency of intelligent production line.In this paper,computer-computing power,data writing processing time and a large amount of data query time are used as platform operation measurement indicators.The experimental results show that the platform can effectively solve the bottleneck problem of massive data processing and analysis of terminal equipment for the client.

关 键 词：大数据 物联网 平台即服务 API Spark框架 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]