基于非线性预测滤波的卫星姿态确定

Satellite Attitude Determination Based on the Nonlinear Predictive Filter

为了提高卫星姿态确定系统的精度和可靠性,采用由姿态动力学与运动学方程构成的非线性卫星姿态运动模型,结合陀螺和矢量观测信息来确定卫星姿态。针对转动惯量变化等引起的模型不准确问题,研究基于预测滤波实现姿态估计的关键技术。仿真结果表明,在恰当选取加权阵的前提下,该方法可实时估计模型误差,当用低通滤波去除高频噪声的模型误差进行补偿后,大大削弱了模型不准确对姿态估计精度的影响,其估计效果明显优于传统的扩展卡尔曼滤波,由此验证了预测滤波在卫星姿态确定中的有效性和精确性。

In order to improve the accuracy and reliability of the satellite attitude determination system, the satellite attitude motion is modeled by the nonlinear attitude dynamics and kinematics equations and combined with gyro and vector observations to determine the satellite attitude. According to the model inaccuracy that caused by the inertia change etc, the key technique of attitude estimate based on the nonlinear predictive filter is studied. The simulation results show that this method can estimate the model error in real-time under the condition of the appropriate weighted matrix. The high frequency noise can be removed through the low pass filter. The impact of model inaccuracy to the accuracy of estimate is greatly weakened. The result of NPF is better than the traditional extended Kalman filter. Accordingly, the validity and accuracy of NPF in satellite attitude determination is verified.