摘要
在对开环干涉型光纤陀螺仪大量实验数据分析的基础上,分别对光纤陀螺仪零位温度漂移建立BP神经网络温度补偿,标度因数与温度、输入角速率建立多项式模型。利用建立的模型对实验数据进行温度补偿。补偿结果表明,BP神经网络补偿效果优于多项式模型,零偏和零偏稳定性减小了一个数量级,补偿效果明显。
Based on analysing lots of experimental data of open-loop interference type fibre optic gyro （FOG） , we set up the BP neural network temperature compensation for zero temperature drift of FOG and the polynomial model for scale factor, temperature and input angular rate respectively. The established model is used to make temperature compensation on experimental data. The results of compensation show that, BP neural network is better than polynomial model in compensation effect, the zero-bias and the stability of zero-bias is reduced by an order of magnitude, the compensation effect is noticeable.
出处
《计算机应用与软件》
CSCD
北大核心
2013年第11期54-56,共3页
Computer Applications and Software
基金
国家自然科学基金项目(61201120)
国家高技术研究发展计划项目(2010aa7010213)
关键词
光纤陀螺仪
温度补偿
BP神经网络
多项式模型
Fibre optic gyro
Temperature compensation
BP neural network
Polynomial model
