计算机工程与应用 ›› 2020, Vol. 56 ›› Issue (16): 273-278.DOI: 10.3778/j.issn.1002-8331.1905-0124

• 工程与应用 • 上一篇    

油田物联网系统边缘计算研究与实践

石锋,梁尚斌,蒋勇,闫丽萍,徐建军,郭晓东   

  1. 1.中国石油化工股份有限公司 西北油田分公司,乌鲁木齐 830011
    2.四川大学 电子信息学院,成都 610065
  • 出版日期:2020-08-15 发布日期:2020-08-11

Research and Practice on Edge Computing of Oilfield Internet of Things System

SHI Feng, LIANG Shangbin, JIANG Yong, YAN Liping, XU Jianjun, GUO Xiaodong   

  1. 1.Northwest Oil Field Company, Sinopec Corp, Urumqi 830011, China
    2.College of Electronics and Information Engineering, Sichuan University, Chengdu 610065, China
  • Online:2020-08-15 Published:2020-08-11

摘要:

随着智能传感器和无线通信技术的发展,油田物联网系统提高了现场生产数据采集的频率和生产过程控制的效率,然而现有物联网系统仍然通过位于远程数据中心的计算资源进行数据处理和控制,网络带宽和通信延迟成为严重的瓶颈。通过对物联网系统的边缘层设备应用边缘计算技术,充分利用边缘网关的计算和存储能力,使用孤立森林算法实现异常数据检测和报警规则学习,同时对温度和阀门开关进行逻辑控制,将之前在云端的处理功能下沉在边缘端实现,降低对网络的要求,满足偏远地区油田生产需要。

关键词: 物联网, 边缘计算, 孤立森林

Abstract:

Owing to the rapid development of intelligent wireless sensor and wireless communication techniques, the Internet of Things(IoT) has considerably improved the frequency of data acquisition and the efficiency of production process control for oil industry applications. However, the existing IoT system still processes the sensed data and issues a control command through the computing resources located in distant data center. As a consequence, network bandwidth and communication latency become serious bottlenecks. Edge computing technology is used to edge intelligence equipment in IoT application, it makes full use of the computing and storage capacity at the edge of the gateway, uses isolation forest algorithm to carry out anomaly data detection and automatically learn alarm rules, and simultaneously realizes the logic control of the temperature and the valve switch of furnace heater. The processing function previously in the cloud is implemented at the edge, so as to reduce the requirements on the network latency and meet the oil industrial application needs in remote areas.

Key words: Internet of things, edge computing, isolation forest