基于模糊RBF神经网络的传感器动态特性补偿研究  

Dynamic Compensation Algorithm for Sensor Based on Fuzzy RBF Neural Network

作　　者：王军号[1] 孟祥瑞[1] 

机构地区：[1]安徽理工大学,安徽淮南232001

出　　处：《煤矿安全》2010年第7期19-22,共4页Safety in Coal Mines

基　　金：安徽高校省级自然科学研究重点资助项目(KJ2010A084)

摘　　要：为了改善传感器的动态特性,减小系统测量误差,分析了传感器动态性能补偿的基本原理,把模糊RBF神经网络应用到传感器的补偿环节。仿真实验表明,使用补偿的传感器输出达到稳态的时间比没有补偿的缩短了大约9ms,相应的动态特性指标也得到了较大的改善。把该算法用于对瓦斯传感器的非线性校正,大大提高了瓦斯检测的灵敏度和精度。To improve sensor＇s dynamic performance,and reduce errors in systematic measurement,the principle of sensor＇s dynamic performance compensation was analyzed,and fuzzy RBF neural network was applied to the sensor＇s compensation. Experiment shows that time for the sensor output to reach steady state with compensation became 9ms shorter than those without compensation,and corresponding dynamic indicators was also largely improved. And the algorithm was used for calibration of the methane detection,which can greatly enhanced the sensitivity and accuracy of gas detection.

关 键 词：动态特性补偿 模糊逻辑 RBF神经网络 瓦斯传感器 

分 类 号：TD605[矿业工程—矿山机电]