一种组合导航系统快速滤波方法及半物理仿真

在利用卡尔曼滤波器对数据进行处理时,协方差矩阵的预报运算过程需要的计算量最大,每一步迭代的计算量与n³(n为状态矢量维数)成正比,约占整个卡尔曼滤波过程70%的CPU时间.协方差阵预报计算过程中,数据输入输出所需要的传递工作量也最大.由于微小型飞行器导航系统采用小体积、低功耗、低成本的微处理器作为导航计算机,为了保证导航实时性的要求,提出了一种降维滤波器加矩阵外积法的快速滤波方法来减少MIMU(Micro Inertial Measurement Unit)/GPS(Global Positioning System)/MMC(Micro Magnetic Compass)组合导航滤波算法的计算量,以提高算法的实时性.半物理仿真试验结果表明:此种算法不仅可以提供较为满意的导航精度,而且大大减小了计算量,提高了系统的实时性.

Fast data fusion method for integrated navigation system and hardware in loop simulation

The main factor in determining the computation time of Kalman filter is the dimension of the model state vector.The number of computations per iteration is direct proportion of n cube.Any reduction in the number of states will significantly decrease the computation time.Since 70 percent of the filtering calculation task is the forecast equation of covariance matrix,the method of reduceddimension model and matrix product is applied in the micro inertial measurement unit/global positioning system/micro magnetic compass(MIMU/GPS/MMC) integrated navigation system,and the hardware in the loop simulation results based on the micro processor prove that the method not only can provide satisfactory accuracy for aircraft navigation but also can reduce the filtering calculation.