基于最大熵卡尔曼滤波的机载SINS星敏感器辅助空中标定技术

为增强机载捷联惯导系统（SINS）在自标定过程中的可观测性，提升陀螺仪漂移和加速度计零偏估计的速度和精度，引入星敏感器姿态信息和GPS速度信息，辅助完成捷联惯导系统的空中标定。同时，考虑在实际空中飞行条件下，受气流、电磁干扰等影响，姿态和速度的量测噪声呈非高斯分布且噪声统计特性不精确，导致经典卡尔曼滤波性能降低。为有效利用量测信号中的高阶矩信息，在卡尔曼滤波中采用最大熵准则代替最小均方误差准则，对星敏感器辅助下的机载捷联惯导系统的误差进行标定。仿真结果表明，经最大熵卡尔曼滤波后，惯性器件误差的标定精度明显提升；在采用星敏感器后，对陀螺仪漂移的标定速度和精度都得到了提升。

Maximum Correntropy Kalman filter based self calibration of Airborne SINS adopting aided by star sensor

Absrtact: In order to improve the observity of velocity and attitude and the eveluate effects of gyro draft and accelerometer error in self calibration of airbone strapdown inertial navigation system, attitude from star sensor and velocity from GPS were introduced to calibrate the strapdown inertial navigation system. Meanwhile the measurment noise of attitude and velocity was affected by pulse noise, becoming non-Gaussian distribution, which conducted performance degeneration of Kalman fiter. In order to utilize high-order moment information of measurement signals, minimum mean square error criterion was replaced by maximum correntropy criterion in Kalman filter, which was used to calibrate the airbone strapdown inertial navigation system aided by star sensor and GPS. The simulation results show that the caliburation accuracy is improved by maximum correntropy Kalman filter. And both of speed and accuracy in calibration were improved by the aide of star sensor.