基于神经网络算法的LVDT传感器非线性补偿方法设计

Design of LVDT Nonlinear Compensation Method Based on Neural Network Algorithm

针对线性可变差动变压器式传感器(Linear Variable Differential Transformer,LVDT)存在非线性缺陷,提出了一种新的级联补偿方法。它是一种基于传感系统的以函数连接型人工神经网络FLANN模型为线性变量的自适应非线性补偿方法,FLANN复杂度不高却拥有高精度的优点。首先分析了LVDT传感器产生非线性的原因,然后利用神经网络算法进行非线性校正,经过实验分析与结果比对,证明了该方法具有较强的可行性、有效性,达到了理想的实验要求。

Considering nonlinear defects of a linear variable differential transformer（ LVDT）,a new cascade compensation method was proposed. It＇s an adaptive nonlinear compensation method which has a sensing system based and has a functionally-linked artificial neural network model（ FLANN） taken as the linear variable.The FLANN has advantages of low complexity and high precision. Analyzing the causes of non-linearity of the LVDT sensor and then using the neural network algorithm to implement nonlinear correction and comparing the results prove both feasibility and effectiveness of this method.