基于小波神经网络技术的井下瓦斯传感器故障诊断分析

煤矿井下瓦斯传感器的准确性和鲁棒性关系到矿井的安全生产。以瓦斯传感器故障诊断为研究对象，利用小波包分解算法分析了瓦斯传感器故障信号、提取了相关特征，并基于这些特征，采用扩展滤波算法训练神经网络。兖州集团王庄煤矿的100组瓦斯检测数据验证了该技术的有效性和准确性，为瓦斯传感器故障的在线诊断系统设计提供了参考。

Analysis of Underground Gas Sensor Fault Diagnosis Based on Wavelet Neural Network

Coal mine methane sensor accuracy and robustness are related to the mine safety production. Taking gas sensor as the research object，fault signal of the gas sensor was analysed based on the fault diagnosis，relevant features were extracted. The extended Kalman filtering algorithm was used to train the neural network based on these obtained features. 100 group gas detection data of Wangzhuang Coal Mine of Yanzhou Group have verified the validity and accuracy of the technique，and have provided a reference for the online diagnosis system for fault of gas sensor.