【摘要】 为了减小半球谐振陀螺在温度变化过程中增益误差、相位误差、非线性误差等通道误差与陀螺零偏的耦合，提升陀螺全温精度，提出了一种基于Elman神经网络对通道误差补偿的方法。首先，分析了增益误差、相位误差、非线性误差等通道误差对陀螺输出的影响机理，其次，建立了基于Elman神经网络的补偿模型。最后通过实验对建立的模型进行了验证。实验结果表明，由于辨识并补偿了耦合在陀螺阻尼不均中的通道误差，零偏在全温过程中的变化量减小，非线性得到改善，陀螺的全温精度由0.1132(°)/h提升至0.012(°)/h。更多还原

【Abstract】 In order to reduce the coupling of the channel errors such as gain error, phase error and nonlinear error with zero offset of the hemispherical resonant gyroscope in the process of temperature change, and improve the accuracy of gyroscope at full temperature, a method of channel error compensation based on Elman neural network is proposed. Firstly, the principles by which channel errors such as gain error, phase error and nonlinear error affect the gyroscope output are analyzed, and secondly, a compensation model based on Elman neural network is established. Finally, the established model is verified by experiments. The experimental results show that due to the identification and compensation of the channel errors coupled in the uneven damping of the gyroscope, the variation of zero offset during the full-temperature process is reduced, the nonlinearity is improved, and the accuracy of the gyroscope at full-temperature is increased from 0.1132(°)/h to 0.012(°)/h.更多还原