摘要
动态条件下分时复用三轴光纤陀螺相位差导致工作轴精度降低,目前尚无较好的解决办法。由分时复用陀螺模型参数推导出系统输入输出的差分方程,进行幅频和相频仿真,得到相位差幅值大小直接影响系统的精度。为此首次提出了基于数据趋势估计的m点迭代相位差估计算法,重点分析了基于数据趋势的估计算法普遍存在的锁死机理,提出了基于调整跳转系数和引入遗忘因子的解决办法,并加入判断准则以抵抗异常值的影响。讨论了算法的收敛速度及其在实际中应用的问题。通过仿真实验验证了算法的可行性和正确性。从实验数据中可以反映出在动态条件下,该算法可以使得分时复用陀螺单轴精度提高1.3倍以上。
The accuracy problem induced by phase error in time-division multiplexing of three-axis fiber optic gyroscope in dynamic situation has not been solved effectively.The system input and output difference equation is derived by the gyro model parameters.By simulating the equation in amplitude and phase-frequency region,the amplitude of phase error has relation to the system accuracy without any condition of phase-frequency impact.M-point iteration phase error estimation algorithm is provided at first time.The steps of the algorithm and its flowchart are made in detail,in which lock-dead mechanism here and in other general data trend estimation algorithms is analyzed.And an effective solving way based on the leap coefficient and a forgetting factor is also provided.A judgment rule is introduced to avoid the impact of exceptional data.Then the convergence rate and the practical application of this algorithm are discussed in several comparison simulations and experiments.A conclusion from the experimental data in dynamic situation is that the algorithm can improve the precision of single axis with time-division multiplexing more than 1.3 times.
出处
《计算机仿真》
CSCD
北大核心
2010年第11期48-52,共5页
Computer Simulation
关键词
相位差估计
锁死
分时复用
光纤陀螺
Phase error estimation
Lock-dead
Time-division multiplexing
Fiber optic gyroscope
