人工神经网络在电涡流传感器非线性补偿中的应用研究  

Application of ANN in Electric Eddy Sensor Nonlinear Compensation

作　　者：张元敏[1] 王红玲[1] 

机构地区：[1]许昌学院电气信息工程学院,河南许昌461000

出　　处：《现代电子技术》2008年第10期190-191,共2页Modern Electronics Technique

摘　　要：电涡流传感器输入/输出特性曲线具有很强的非线性,为保证仪表输入/输出的线性化关系,必须对电涡流传感进行非线性补偿。采用RBF网络对电涡流传感器进行非线性补偿,并将补偿结果与拟合函数法(最小二乘法)进行比较。结果表明:神经网络产生的补偿曲线更顺滑,预测性更强,补偿后的传感器线性度更好。characteristic curve of the eddy current sensor is serious nonlinear. The nonlinear errors of the eddy current sen- sor must be compensated in order to ensure the instrumen（s linearity. The RBF network is adopted and the result is compared with that from the fitting function method. The comparative result indicates that the compensation curve of adopting BRF net- work is more smoother,the forecast character is more higher and the sensorrs linearity is better.

关 键 词：RBF神经网络 电涡流传感器 非线性补偿 拟合函数 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]