逆协方差交叉融合鲁棒Kalman滤波器

Inverse Covariance Intersection Fusion Robust Steady-State Kalman Filter

分布式状态估计系统通过将多个传感器状态融合以得到更精确的融合结果,当传感器之间的协方差未知时,常采用保守估计的策略,但结果精确度较差。为了在传感器之间互协方差未知时得到更精确的融合结果,引入了逆协方差交叉算法,将其与局部稳态Kalman滤波器相结合,提出逆协方差交叉融合鲁棒Kalman滤波器。它克服了协方差交叉融合(CI)算法保守的缺点,证明了ICI的精度高于CI的精度,并基于协方差椭圆给出ICI、CI和局部传感器精度的几何解释。通过两传感器系统的蒙特卡洛仿真例子表明,其实际精度相比于CI融合鲁棒稳态Kalman滤波器更接近于带已知互协方差的最优融合器的精度。

Aimed at the problems that in distributed state estimation systems,the fusion methods are often employed to systematically combine multiple estimates of the state into a single,more accurate estimate,and if the correlation structure is unknown,conservative strategies are typically pursued with less accurate,an inverse covariance intersection fusion robust steady-state Kalman filter is proposed to gain more accurate estimate.As a major advantage of the novel approach,the fusion results prove to be more accurate than those provided by the well-known covariance intersection method.The geometric interpretation of the accuracy relations is given based on the covariance ellipses.A Monte-Carlo simulation example for a two-sensor system shows that its actual accuracy is close to that of the optimal Kalman fuser with known cross-covariance.