一种高精度机载POS双位置对准方法

Two-position alignment method of airborne position and orientation system

针对惯性器件输出噪声引起高精度机载POS(Position and Orientation System)地面双位置对准精度较差的问题,提出基于小波滤波和隐马尔科夫建模的数据预处理方法结合自适应卡尔曼滤波的双位置对准方法。首先分析惯性敏感器原始信息的频率特性,利用小波滤波算法,消除惯性器件测量中的高频噪声;综合分析器件的随机游走特性,通过建立隐马尔科夫模型削弱惯性敏感器输出随机游走的影响;并针对降噪处理、电源波动及环境因素等引起的系统噪声统计规律不确定性问题,提出利用自适应卡尔曼滤波的方法实现POS高精度初始对准。试验结果表明,采用本文所提方法的对准结果,可使对准结束后600s纯捷联解算的水平速度误差由1.278 m/s减小至0.6061 m/s,水平位置误差由274.6m减小至128.2 m,水平速度和位置误差均减小了50%左右。

In view of the bad precision caused by inertial sensor noise for airborne POS（Position and Orientation System）, a preprocessing method based on wavelet filter and HMM（Hidden Markov Model） combined with AKF（Adaptive Kalman Filter） is proposed. Firstly, the spectrum characteristic of the inertial sensors is analyzed, and a wavelet filter is employed to eliminate the high-frequency noise. Then the HMM is build to decrease the influence of random walk error. The inertial information is processed, so the noise statistical distribution of the system is uncertain. An adaptive Kalman filter is used to improve alignment accuracy. Experiment results indicate that, after 600 s strapdown calculation with the proposed method, the horizontal velocity error decreases from 1.27 m/s to 0.6061 m/s, and the horizontal position error decreases from 274.6 m to 128.2 m, and horizontal velocity and position error are reduced by about 50%.