预测滤波算法在微小卫星姿态确定中的应用

Application of Predictive Filter in Satellite Attitude Determination

讨论了利用卫星姿态动力学与矢量观测信息融合确定卫星姿态的方法。引入基于模型误差最小准则的预测滤波算法,实现对未知扰动的实时辨识与估计,克服模型误差影响。针对预测滤波算法存在模型误差加权阵依赖经验选择的不足,提出了模型误差加权阵偏小设计准则,使预测滤波算法对模型误差的估计引入高频分量,根据卫星姿态机动速度远小于高频噪声的事实,采用FIR数字滤波器滤除由高频分量引入的噪声,放宽了模型误差加权阵的设计范围。仿真表明:应用该设计准则的预测滤波算法与优化模型误差加权阵的估计效果相当,且滤波精度对模型误差加权阵参数的选择不敏感;该法也克服了优化模型误差加权阵在滤波初始时刻的振荡。

Satellite attitude determination method using attitude dynamics and vector observations is discussed.Predictive filtering based on minimum model error is introduced to realize real time identification of unknown perturbations and conquer influence of model error.The selection of model error weight matrix(MEWM) of predictive filter depends on experience,so lower design criterion of MEWM is presented in the paper,then high frequency noise will be introduced in the estimation of model error.According to the fa...