基于提升小波与灰色神经网络的光纤陀螺振动误差建模

摘要

FOG sensors have large random noises and drifts under vibration environment. In order to compensate them by using accurate algorithm, a proper vibration error model must be established. In this paper, the output signal of digital close-loop FOG under vibration is analyzed using Allan variance method. Then the white noises and drift errors in the error model of FOG are separated by using lifting wavelet, and a compound modeling method for drift errors of FOG is put forward based on the grey theory and RBF neural network. The simulation results show that, compared with traditional RBF neural network, the proposed modeling method can effectively filter out the effects of white noise, and the modeling precision of drift model for FOG sensor is approximately doubled. The output precision of FOG can be improved effectively by using the proposed method, which is useful to the study of FOG performance under vibration environment%光纤陀螺在振动环境下的输出具有噪声大、漂移强的特性,必须建立合理的振动误差模型,以便使用精确的算法进行补偿,从而提高光纤陀螺的输出精度.文中首先使用Allan方差分析法分析了某型号的数字闭环光纤陀螺在振动环境下的输出信号,随后利用提升小波分离出了光纤陀螺误差模型中的白噪声及漂移误差,并提出了基于灰色理论和RBF神经网络的漂移误差建模方法.仿真结果表明,相较于传统的RBF神经网络模型,基于提升小波的灰色RBF神经网络的漂移误差建模方法能有效滤除白噪声,并将漂移误差模型的建模精度提高了一倍左右.该方法能够有效提高光纤陀螺在振动环境下的输出精度,对光纤陀螺在振动环境下的误差研究具有重要指导意义.