基于自适应卡尔曼滤波的光纤陀螺噪声系数估计方法

针对Allan方差法确定光纤陀螺ARW（angle random walk）噪声系数的一些不足,如大量存储数据、非实时处理、计算量大、耗时长等,提出了基于自适应卡尔曼滤波的光纤陀螺ARW系数在线估计方法.在角度随机游走、零偏不稳定性、角速率随机游走等主要噪声数学特性分析基础上,建立了光纤陀螺现代状态空间噪声误差模型,基于新息自适应卡尔曼滤波量测噪声协方差阵的迭代计算,实现光纤陀螺ARW系数的在线、实时估计,从而避免了存储大量历史数据,显著地减小了计算量,缩短了陀螺数据处理时间.数字仿真试验和光纤陀螺实测数据试验结果均验证了本文方法的可行性和有效性.

Optical fiber gyro noise coefficient estimation method using adaptive Kalman filtering

To overcome some shortcomings of Allan variance method while determining optical fiber gyro ARW （angle random walk） noise coefficient,such as need storage of amount of data,non real-time processing,huge computational burden,time consuming,an online estimation method for fiber optic gyro ARW coefficient based on adaptive filtering is proposed.Based on analysis of the mathematics characteristics of angle random walk,bias instability,rate random walk noise,the modern state space error model for fiber optic gyro is established.Online,real-time estimation of fiber optic gyroscope ARW coefficient is realized through iterative calculation of measurement noise covariance matrix in innovation based adaptive Kalman filtering,so as to avoid the storage of large amounts of history data,greatly reduce the computational burden,shorten the time for processing gyro data.Test results of both digital simulation and measured data of fiber optic gyros verify the validity and feasibility of the proposed approach.