论文部分内容阅读
对于带未知互协方差的两传感器系统,提出一种协方差交叉(CI)融合鲁棒稳态Kalman滤波器,它关于未知互协方差具有鲁棒性.严格证明了该滤波器的实际精度高于每个局部滤波器的精度,但低于带已知互协方差的最优融合Kalman滤波器的精度.基于协方差椭圆给出了精度关系的几何解释.进一步将上述结果推广到一般多传感器情形.一个跟踪系统的Monte-Carlo仿真例子表明,其实际精度接近于带已知互协方差的最优融合器的精度.
For a two-sensor system with unknown cross-covariance, a covariance cross (CI) fusion robust steady-state Kalman filter is proposed, which is robust to unknown cross-covariance and strictly verifies the high precision of the filter The accuracy of each local filter is lower than that of the optimal fused Kalman filter with known mutual covariance. The geometric interpretation of the accuracy relationship is given based on the covariance ellipse. The above results are generalized to the general multi-sensor The Monte-Carlo simulation example of a tracking system shows that its actual accuracy is close to that of the optimal cofactor with known mutual covariance.