Study Of Robust Filtering Approach For INS/BD Integrated Navigation System

Inertial navigation system with Beidou navigation system may constitute a combination of high precision navigation system with complementary advantages, as the pinnacle of integrated navigation researching field, the technology of data fusions closely related to the degree of our country’s independent satellite navigation technology. This article mainly around the study of robust filtering technology of the inertial/beidou integrated navigation,and the research contents would be divided into several aspects:Firstly, we introduce the working principle and the error model of the navigation subsystem, and kinds of combination of subsystems as well, then select the theoretical basis on loose combination model and establish the mathematical model about the system with the principle of data fusion. Meanwhile, we simplify the state variables of the system and reduce the order of the model combining with the characteristics of the robust filtering algorithm,using the high-order treatment instead of the usual method so as to reduce the model error introduced by the process of the first-order approximation on some extent, and it is more suitable to propose a new solution for the limitations and the drawbacks of conventional algorithm son data fusion.Next, we research the robust filtering algorithm based on the inertial/beidou integrated navigation system of continuous-time, with the simplified mathematical model as the reference model, the robust filters are designed to meet the different performance index and to insure the error system is asymptotically stable, the technique of linear matrix inequality(LMI)is used here to solve the feasible solution for the filters. Numerical simulation can verify the advantages of robust filtering under the condition of model disturbances, the robust filtering algorithm can effectively inhibit the outside noise interference and improve the observability and accuracy of the system.Finally, based on the study of robust filtering technology in continuous-time system, and considered the realizability of practical engineering, we research the robust filtering algorithm of discrete-time system, using the discrete model as the research object, robust filters are designed to meet different performance index, and we use the combination of hardware-in-the-loop simulation and off-line simulation calculation to verify the validity of these algorithms. The simulation results show that the robust filtering algorithm can apply in practical inertial/beidou integrated navigation system effectively, and reduce the effect ofinterference on the filtering accuracy, improve the accuracy and reliability of the filtering. At the same time, the research contents for data fusion technology of the integrated navigation system put forward new research direction, and have certain practical significance and research value.