Abstract: In relative navigation model, there are linear model and nonlinear model. And the navigation algorithm based on Kalman filtering has high computational complexity in multisensor information fusion. Aiming at these problems, the hybrid information filtering is proposed. To overcome the influence of inaccuracy of noise statistic characteristic, the relative navigation based on adaptive hybrid information filtering is also proposed. Theoretical analysis and numerical simulation show that the hybrid information filtering has lower computational complexity in multisensor information fusion than the navigation algorithm based on Kalman filtering. And the navigation task with linear and nonlinear hybrid models also can be completed with the hybrid information filtering. Besides the above characteristics, via adaptive adjusting the measurement covariance matrix, the effect of inaccurate noise statistic characteristic can be reduced.

Key words: multisensor information fusion, information filtering, relative navigation, adaptive navigation