基于挠性陀螺仪动力学模型,研究参数辨识在陀螺仪性能分析中的应用方法。在论述挠性陀螺仪性能指标和力学模型的基础上,利用伪随机多谐波信号作为激励信号,通过零阶保持器将离散模型转换为连续模型进行参数辨识。实验结果显示,模型预测输出与实际测量数据的拟合度达到95.3%,残差在合理的公差带以内。因此,参数辨识是陀螺仪性能分析中的一个有效性方法,可为陀螺仪故障分析和控制回路设计提供一种技术手段和理论依据。
The paper presented a performance analysis method for flexible gyroscope. This method was based on the key per- formanee indicators and the dynamic model of flexible gyroscope and it was applied in the following steps. Firstly, muhisine, which was also called pseudorandom fioise, was chosen as the excitation signal. With the excitation signal and response signal, and applying special algorithm, the preliminary discrete model could be obtained. Secondly, converted this discrete model to a continuous model using zero-order-hold method, and applied Gauss-Newton minimization method to refine the initial estimation to get the optimal model. The experimental results indicates that the fitness degree increases to 95.3% and the residual error is in the tolerance band. The proposition, which is illustrated with practical results, provides a feasible technical measure and a theoretical basis.