基于SVR的瓦斯传感器故障诊断方法  

Method of gas sensor fault detection based on SVR

作　　者：付华[1] 杨欣[1] 高婷[1] 

机构地区：[1]辽宁工程技术大学电气与控制工程学院,辽宁葫芦岛125105

出　　处：《传感器与微系统》2009年第2期10-12,共3页Transducer and Microsystem Technologies

基　　金：辽宁省科技攻关计划资助项目(2007231003);辽宁省优秀人才基金资助项目(2007R24)

摘　　要：介绍了回归型支持向量机(SVR)的基本原理,建立了基于SVR的传感器时间预测模型,对时间预测器实行离线训练、在线应用的方法,用训练好的SVR模型模拟煤矿井下瓦斯传感器系统的动态特性,阐述了瓦斯传感器故障诊断和信号恢复的实现过程。仿真结果表明:SVR时间预测器能准确预测和跟踪瓦斯传感器的输出,及时诊断出传感器故障信息,并对传感器信号进行恢复,实验验证了该方法的正确性和有效性。The theory of support vector regression （SVR） is introduced, a sensor predictor based on SVR is proposed. The SVR is trained out of line and used on line. The trained SVR model is use to simulate the dynamic characteristic of coal mine gas sensor system. The implementation procedures of gas sensor fault detection and signal recovery are expounded . The simulation results show that the SVR predictor can predict and follow the output of gas sensor accurately, fault message of sensor is diagnosed in time, and sensor signal is recoved. The method is verified by the experiment to be correct and valid.

关 键 词：回归型支持向量机 瓦斯传感器 故障诊断 信号恢复 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]