欢迎访问《金属矿山》杂志官方网站,今天是 分享到:
×

扫码分享

金属矿山 ›› 2014, Vol. 43 ›› Issue (01): 120-124.

• 安全与环保 • 上一篇    下一篇

基于无线ZigBee的钨矿尾矿库安全监测系统

何学文,盛颖飞,曹清梅   

  1. 江西理工大学机电工程学院,江西 赣州 341000
  • 出版日期:2014-02-14 发布日期:2014-02-25
  • 基金资助:

    国家自然科学基金项目(编号:61163063,50764005),江西省教育厅科技项目(编号:GJJ12329,GJJ12353)。

Tungsten Tailings Pond Safety Monitoring System based on Wireless ZigBee

He Xuewen,Sheng Yingfei,Cao Qingmei   

  1. Mechanical and Electrical Engineering College,Jiangxi University of Science and Technology,Ganzhou 341000,China
  • Online:2014-02-14 Published:2014-02-25

摘要: 针对当前钨矿尾矿库安全监测系统存在安装和维护成本高、布线复杂、传输距离及能量受限等问题,设计了一种基于ZigBee无线传感器网络和LabVIEW的监测预警系统。选用片上芯片CC2530作为射频收发器,完成了太阳能供电的传感器节点的硬件和软件设计,实现对尾矿库监测数据的采集、传输和处理。上位机使用LabVIEW设计监测预警界面,实时显示监测参数并及时预警,利用VISA串口资源模块和SQL语言调用的Access数据库,实现监测数据的传输和存储。上位机数据处理模块通过LabVIEW的Matlab Script节点在后面板中编程调用Matlab软件并建立回归型支持向量机(SVR)模型进行测试。实验结果表明,系统能实时采集监测参数,SVR回归模型预测误差为0.3%左右,适合钨矿尾矿库的参数预测,对控制钨矿尾矿库风险,确保其安全意义重大。

关键词: 无线传感器网络, 钨矿尾矿库, ZigBee, LabVIEW, SVR

Abstract: In view of the existing problems of high installation and maintenance cost,complex wiring,limited transmission distance and energy in current safety monitoring systems of tungsten tailings pond,a new type of monitoring and warning system based on ZigBee wireless sensor network and LabVIEW is designed.The system chooses system-on-chip CC2530 as the radio frequency transceiver and completes hardware and software designing of the solar power sensor nodes,which could collect,transmit and process the data of tailings pond.The monitoring and warning interface of PC is completed with LabVIEW,which displays real-time monitoring parameters and warns in time.It uses VISA serial port resource module and Access database called by SQL to transmit and store the monitoring data.Through the Matlab Script node,data processing module of PC calls Matlab software and establishes the regression model of support vector machine (SVR) in the back panel of the LabVIEW.The experimental results show that the system could collect real-time monitoring parameters.The prediction error of SVR regression model is about 0.3%.It is suitable for predicting parameters of tungsten tailings pond.The system is of great significance to control the risk and ensure the security of tungsten tailings pond.

Key words: Wireless sensor networks, Tungsten tailings pond, ZigBee, LabVIEW, SVR