分析煤矿安全监测系统中瓦斯传感器的特性,探讨其在线使用中产生的主要故障类型及模式。介绍了一种基于人工神经网络进行瓦斯传感器故障检测的新方法,利用单个瓦斯传感器的输出信息为瓦斯传感器建立了动态非线性神经网络传感器输出模型,并利用该模型进行在线故障检测。实验证明该模型具有良好的收敛性和稳定性,能满足对瓦斯传感器故障在线检测的需要。
The paper analyzed the features of the gas sensors used in the mine safety monitoring and measuring system, The paper discussed the main failure type and mode occurred in the on-line application, The paper introduced a new method of the gas sensor failure detect/on base on an artificial nervous network. An output mode of the dynamic non-linerar nervous network sensor was established for the gas sensor base on the output information of a single gas sensor. The mode was applied to measure and detect the on-line failure. The test results showed that the mode had a good convergence and stability, which could fully meet the requirements of the on-line measurement and detection of the gas sensor failure.