离散制造车间实时定位数据可视化平台设计  

Design of Real-Time Location Data Visualization Platform for DiscreteManufacturingWorkshops

作　　者：耿向征 袁逸萍[1] 毛军 吴正春 GENG Xiang-zheng;YUAN Yi-ping;MAO Jun;WU Zheng-chun(School of Mechanical Engineering,Xinjiang University,Xinjiang Urumqi 830047,China;Zhuolang Xinjiang Intelligent Machinery Co.,Ltd.,Xinjiang Urumqi 830000,China)

机构地区：[1]新疆大学机械工程学院,新疆乌鲁木齐830047 [2]卓郎新疆智能机械有限公司,新疆乌鲁木齐830000

出　　处：《机械设计与制造》2023年第7期236-241,共6页Machinery Design & Manufacture

基　　金：自治区科技支疆项目(2019E0213);自治区重点研发(202008748)。

摘　　要：室内实时定位技术被广泛地应用于获取离散制造车间中生产要素的位置信息。然而,传统的单机计算方法已不能满足制造大数据环境下定位数据的实时处理需求。通过利用UWB实时定位技术、Apache Kafka、Spark Streaming、HBase等大数据技术及Demo 3D软件,搭建一套制造车间定位数据获取、收集、计算、存储和可视化的平台,并设计了基于Spark流式处理框架下的多线程卡尔曼滤波器,相较于传统单机多线程计算方式快218.2s。实现了对车间中叉车、人员、关键物料的实时定位、监管和海量数据的存储,提高了车间的透明化管理水平。Indoor real-time location technology is widely used to obtain essential productive factors'location information in discrete manufacturing workshops.However,the traditional single-machine computing method cannot meet the needs of real-time processing of positioning data in the manufacturing environment of big data.By using UWB real-time positioning technology,big data technology include Apache kafka,Spark Streaming HBase and Demo 3D software,set up a manufacturing plant location data capture,collection,calculation,storage and visualization platform,and designed a framework based on Spark Streaming processing of multi-threaded kalman filter,218.2 seconds faster than traditional single multithreaded computing paradigm.The real-time location,supervision and storage of massive data of forklifts,personnel and key materials in the workshop were realized,and the transparent management level of the workshop was improved.

关 键 词：离散制造车间 大数据技术 卡尔曼滤波器 Demo 3D 可视化 

分 类 号：TH16[机械工程—机械制造及自动化] TP39[自动化与计算机技术—计算机应用技术]