摘要
为了提高MEMS陀螺仪的精度,利用基于灰色累加累减操作能够减小MEMS陀螺仪随机性的特点,提出一种将灰色理论与径向基函数(RBF)神经网络相结合的MEMS陀螺随机误差建模补偿方案:采用Allan方差分析法对MEMS陀螺输出数据构成的样本空间进行处理并辨识信号中的随机误差项;通过灰色累加累减过程与RBF神经网络的嵌入式耦合,实现MEMS陀螺随机误差预测模型的建立。实验结果表明,与MEMS陀螺实测数据比对后可发现灰色RBF神经网络方法能够有效预测多种随机误差,可进一步提高MEMS陀螺仪输出的预测精度。
In order to improve the accuracy of MEMS gyroscope,considering the characteristics that the grey cumulative and regressive operation could reduce the randomness of MEMS gyroscope,the paper proposed a MEMS random modeling method which combines the grey theory and radial basis function（RBF）：firstly,the random errors of the sample space composed of the gyro output data were analyzed with Allan variance;secondly,the random prediction model of MEMS gyroscope was built through the embeded coupling of grey cumulative and regressive process and RBF neural networks.Experimental result showed that the proposed method could effectively predict the random errors and improve the predicted output accuracy of MEMS gyroscope by comparing with measured data.
出处
《导航定位学报》
CSCD
2017年第3期9-13,共5页
Journal of Navigation and Positioning
基金
辽宁省高等学校杰出青年学者成长计划(LJQ2015044)
辽宁省自然科学基金(2015020078)
辽宁省"百千万人才工程"培养经费资助(辽百千万立项【2015】76号)
关键词
MEMS陀螺
灰色模型
RBF神经网络
随机误差
MEMS gyroscope
grey model
radial basis function neural networks
random error
