一种高动态条件下惯性星光组合定姿方法

针对高动态条件下星图拖尾导致惯性星光组合定姿精度下降甚至无法定姿的问题,提出了一种基于乘性扩展卡尔曼滤波的惯性星光深度组合姿态确定方法.利用星敏感器观测信息修正姿态误差,补偿陀螺漂移,并建立了陀螺输出的角度变化量与星图像移的关系,利用陀螺输出信息估计星图拖尾的模糊参数,采用维纳滤波复原方法对产生拖尾的星图进行复原以提高星图信噪比和观测精度.仿真结果表明该方法可以有效提高星像质心提取精度和星图识别率,对初始姿态误差修正更快,且不存在星图误匹配的情况,保证了惯性星光组合定姿方法在高动态条件下仍能保持较高的精度.

A method of stellar-inertial attitude determination under high dynamic condition

Under high dynamic condition, the accuracy of stellar-inertial attitude determination degraded or even failed to obtain the attitude information because of star image smearing. Based on multiplicative extended Kalman filter (MEKF), a deep integration attitude determination method using gyro and star sensor was proposed. The attitude estimation errors and gyro drifts were compensated by using the star sensor information. The relationship between the gyro output and the pixel motion of the image was established, then the smearing parameters could be estimated by using the gyro output. The star image was restored by using Wiener filter with estimated smearing parameters to improve the signal-to-noise ratio and observation accuracy. Simulation results show that the estimation accuracy of star centroid and the recognition rate of star image can be improved remarkably, the initial attitude estimation errors can be modified faster. The method eliminates recognition mistakes and ensures high accuracy of attitude determination under high dynamic condition.