论文部分内容阅读
本文研究了动力调谐陀螺仪全系统辨识方法,针对系统辨识过程中模型复杂度高和传统辨识方法的有色噪声敏感性问题,对全系统模型进行简化分析和降阶处理,获取辨识模型集,并提出正交Levy辨识算法,通过数据矩阵的正交化处理,抑制有色噪声干扰,提高辨识精度.最后,结合某型号动力调谐陀螺仪进行全系统辨识实验.实验结果表明,正交Levy法的辨识拟合度超过90%,双轴连续实验也表明正交Levy法辨识结果稳定可靠,能够应用于动力调谐陀螺仪全系统实时监测.
In this paper, the system-wide identification of dynamic tuned gyroscopes is studied. Aiming at the problem of high sensitivity of the model in the process of system identification and the traditional method of identification, the system-wide model is analyzed and reduced in order to obtain the identification model set. An orthogonal Levy algorithm is proposed to reduce the interference of colored noise and improve the identification accuracy by orthogonalizing the data matrix.Finally, a system-wide dynamic tuning gyroscope is used to perform system-wide identification experiments.The experimental results show that the identification of orthogonal Levy’s method The fitting degree is more than 90%. The biaxial continuous experiments also show that the identification results of quadrature Levy method are stable and reliable, and can be applied to the whole system real-time monitoring of the power tuned gyroscope.