Research On Multi-Sensor Integrated Navigation Algorithm Based On Federated Filter

With the rapid development of information technology,the demand for high precision and high reliability navigation and positioning is also increasing.Since any single navigation system has inherent limitations,it is difficult to meet the needs of modern technology on the performance of navigation systems.As the integrated navigation system has the advantages of complementing each other and giving full play to each sub-navigation system,the research of integrated navigation and positioning based on multi-source sensors has become a research and development hotspot in the field of global navigation system.In this paper,the inertial navigation system is used as the reference system.Based on the information fusion theory,the key technologies of multi-sensor integrated navigation system based on federated filter structure are studied,using multi-sensor navigation data to eliminate the SINS errors and to improve the positioning performance of the system.The main contents of this paper include:(1)The loose SINS/GNSS/DVL/Barometer integrated navigation model under the navigation frame of east,north and upward is constructed.We also present an adaptive federated strong tracking Kalman filter(AFSTKF)algorithm for the purpose of improving the positioning accuracy of the integrated navigation system.In the local filter,STF algorithm with strong robustness is used to improve the accuracy of filtering estimation.In the fusion process of the main filter,a simplified data fusion method is proposed,which can effectively solve the singular problem of covariance matrix,and also reduce the computational complexity.We proposed an adaptive information distribution method based on the statistical characteristic of residual vectors,which can adaptively balance the contributions of local filters.The simulation results show that the proposed AFSTKF algorithm has higher positioning accuracy and more stable performance.(2)Considering that the fault information affects the performance of navigation system,the main factors affecting system fault tolerance in federated filter are studied.It is found that the fault information could affect the fusion result,the fault-free sub-filter will be affected by the feedback process,and the information distribution factor is the main factor influencing the fault tolerance of the sub-filters.In order to improve the reliability and fault-tolerant ability of multi-sensor information fusion system,a two-stage fault-tolerant filtering algorithm based on multiple fading factors is designed.The Chi-square test was used to judge the current fault condition of sub-filters The multi-fading factor adaptive filtering algorithm is used to estimate the state of the sub-filter,which is helpful to enhance the convergence of the filter.In the fusion process of main filter,the estimation information of faulty sub-filter will be isolated,and only the estimation results of fault-free sub-filter are fused to improve the fault tolerance of fusion results.In the feedback process,a fault-tolerant information distribution method is designed to reduce the impact of fault information and enhance the correction effect of global estimation results on fault sub-filter.Finally,the performance of the proposed fault-tolerant algorithm is verified by simulation.