基于SVM的光纤陀螺标度因数建模研究

Model Identification Method Based on SVM for FOG

针对光纤陀螺(FOG)标度因数随温度呈非线性变化的特性,提出采用支持向量机对标度因数进行建模的方法,以减小光纤陀螺输出误差。利用实验数据对支持向量机对进行训练,获得标度因数的温度模型从而提高FOG的精度。同时将支持向量机对标度因数进行建模的结果与传统的最小二乘建模结果进行比较,验证采用支持向量机对标度因数建模是非常有效的。

In view of the non-linearity of the Fiber Optic Gyroscope(FOG) with the change of temperature,a method of modeling for the temperature of scale factor by Support Vector Machine(SVM) is proposed,and it can reduce the output error of FOG.SVM is trained by measured data and thus the parameters of the model can be obtained.The accuracy of FOG is improved by using the model of scale factor.Meanwhile the modeling results of SVM for scale factor are compared with the traditional modeling results,it demonstrates th...