神经网络与组态监控在煤矿瓦斯监测中的应用  

Application of Configuration Control and Neural Network in Coal Mine Gas Detection

作　　者：刘会景[1] 李琨[2] 

机构地区：[1]乌鲁木齐职业大学,新疆乌鲁木齐830002 [2]昆明理工大学,云南昆明650051

出　　处：《煤炭工程》2014年第10期226-228,共3页Coal Engineering

基　　金：云南省教育厅科学研究基金项目(042140D)

摘　　要：针对煤矿井下环境恶劣而引起瓦斯传感器误报警等问题,即瓦斯"大数"问题,研究将组态监控软件和BP神经网路应用于煤矿瓦斯监控系统中,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,不仅可以把干扰消除,同时还能辨别出由于瓦斯突出造成的数据超出设定的百分比浓度点,做到可以准确的报警。Aiming at the gas sensor false alarm caused by poor environment conditions in underground coal mine, configuration monitoring software and BP neural network was applied in coal mine gas monitoring system ; with configuration monitoring software, realizing real - time observation of gas concentration, sensor output checking of any time, display of gas concentration change; and with the design and trainings of BP neural network, the network parameters were constantly optimized, till obtaining the reasonable neural network; the network was tested, to eliminate the interference and distinguish the false alarm data points caused by gas outburst, realizing accurate alarm.

关 键 词：煤矿安全 脉冲干扰 瓦斯浓度 误报警 组态监控 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]