煤炭工程 ›› 2014, Vol. 46 ›› Issue (10): 226-228.doi: 10.11799/ce201410065

• 信息工程 • 上一篇    下一篇

神经网络与组态监控在煤矿瓦斯检测中的应用

刘会景   

  1. 昆明理工大学
  • 收稿日期:2014-07-03 修回日期:2014-08-28 出版日期:2014-10-10 发布日期:2014-10-10
  • 通讯作者: 刘会景 E-mail:lesuscn@126.com

组态监控与神经网络在煤矿瓦斯检测中的应用

  • Received:2014-07-03 Revised:2014-08-28 Online:2014-10-10 Published:2014-10-10

摘要:

针对煤矿井下环境恶劣而引起瓦斯传感器误报警,即瓦斯“大数”问题,本文提出了组态监控软件和BP神经网路在煤矿瓦斯监控系统中的应用,利用组态监控软件可以实时观测瓦斯浓度,查询传感器任何时间的输出,显示瓦斯浓度变化,通过BP神经网络的设计和多次训练,不断修正网络参数,最终确定了合理的神经网络,并对该网络进行测试,该网络有效消除瓦斯浓度中的脉冲干扰引起的误报警现象和防止瓦斯浓度超限时的漏报警。在煤矿安全监控系统中,应用神经网络去除误报警现象具有较好的发展前景。

关键词: 组态监控, 误报警, 脉冲干扰, 瓦斯浓度, 煤矿安全

Abstract:

According to the coal mine environment caused by the bad gas sensor error alarm, namely gas "large" problem, this paper presents the configuration monitoring software and the application of BP neural network in coal mine gas monitoring system, configuration monitoring software can be used to real-time observation of gas concentration, query any time the output of the sensor, according to gas concentration change, through the design of BP neural network and training for many times, and correct network parameters, finally the reasonable neural network, and testing the network, the network effectively eliminate the pulse interference phenomenon of false alarm caused by gas concentration and prevent gas leakage alarm when concentration overrun. In the coal mine safety monitoring system, the application of neural network to remove false alarm has good prospects for development.